Vai ai contenuti. | Spostati sulla navigazione | Spostati sulla ricerca | Vai al menu | Contatti | Accessibilità

| Crea un account

Nan, Fany (2009) Forecasting next-day electricity prices: from different models to combination. [Tesi di dottorato]

Full text disponibile come:

[img]
Anteprima
Documento PDF
2493Kb

Abstract (inglese)

As a result of deregulation of most power markets around the world electricity price modeling and forecasting have obtained increasing importance in recent years. Large number of models has been studied on a wide range of power markets, from linear time series and multivariate regression models to more complex non linear models with jumps, but results are mixing and there is no single model that provides convincing superior performance in forecasting spot prices. This study considers whether combination forecasts of spot electricity prices are statistically superior to a wide range of single model based forecasts. To this end we focus on one-day ahead forecasting of half-hourly spot data from the British UK Power Exchange electricity market. In this work we focus on modeling data corresponding to some load periods of the day in order to evaluate the forecasting performance of prices representative of different moment of the day. Several forecasting models for power spot prices are estimated on the basis of expanding and/or rolling estimation windows of different sizes. Included are linear ARMAX models, different specifications of multiple regression models, non linear Markov switching regression models and time-varying parameter regression models. One-day ahead forecasts are obtained for each model and evaluated according to different statistical criteria as prediction error statistics and the Diebold and Mariano test for equal predictive accuracy. Forecasting results highlight that no model globally outperforms the others: differences in forecasting accuracy depend on several factors, such as model specification, sample realization and forecasting period. Since different forecasting models seem to capture different features of spot price dynamics, we propose a forecasting approach based on the combination of forecasts. This approach has been useful to improve forecasting accuracy in several empirical situations, but it is novel in the spot electricity price forecasting context. In this work different strategies have been employed to construct combination forecasts. The simplest approach is an equally weighted combination of the forecasts. An alternative is the use of adaptive forecast combination procedures, which allows for time-varying combination coefficients. Methods from Bates & Granger (1969) are considered. Models entering the combination are chosen for each forecasting season using the model confidence set method (MCS) described in Hansen et al. (2003, 2005) and then screened with the forecasts encompassing method of Fair & Shiller (1990). For each load period, our findings underline that models behave differently in each season. For this reason we propose a combination applied at a seasonal level. In this thesis some promising results in this direction are presented. The combination results are compared with the best results obtained from the single models in each forecasting period and for different prediction error statistics. Our findings illustrate the usefulness of the procedure, showing that combining forecasts at a seasonal level have the potential to produce predictions of superior or equal accuracy relative to the individual forecasts.

Abstract (italiano)

Con la liberalizzazione dei mercati dell’elettricità, il problema della modellazione e previsione dei prezzi elettrici è diventato di fondamentale importanza. In letteratura sono stati studiati e applicati ad un gran numero di mercati molti tipi di modelli, come modelli per serie storiche, regressione lineare e modelli non lineari a salti molto più complessi. I risultati però sono contrastanti e finora nessun modello ha mostrato una capacità previsiva dei prezzi elettrici superiore rispetto agli altri. L’obiettivo di questa tesi è capire se i modelli di combinazione di previsioni possano dare risultati statisticamente superiori rispetto alle previsioni ottenute da singoli modelli. In particolare, viene affrontato il problema della previsione dei prezzi elettrici del giorno dopo applicato al mercato elettrico britannico UK Power Exchange. In questo mercato, i prezzi hanno frequenza semioraria: al fine di valutare il comportamento previsivo dei modelli, relativamente all’andamento dei prezzi nei diversi momenti della giornata, sono state scelte specifiche fasce orarie. I modelli usati per la previsione dei prezzi sono stati stimati sulla base di finestre di dati espandibili e/o mobili di diverse misure fissate. I modelli considerati includono modelli lineari di tipo ARMAX e diverse specificazioni di modelli di regressione multipla. Inotre sono stati considerati modelli di regressione non lineare a regimi Markov switching e modelli di regressione a parametri non costanti. Le previsioni a un passo ottenute dai modelli specificati sono state confrontate secondo diversi criteri statistici come le statistiche basate sull’errore di previsione e il test di Diebold e Mariano. Dai risultati emerge che, globalmente, nessun modello considerato supera gli altri per abilità previsiva: vari fattori, tra cui specificazione del modello, realizzazione campionaria e periodo di previsione, influenzano l’accuratezza previsiva. Dal momento che modelli di previsione diversi sembrano evidenziare caratteristiche diverse della dinamica dei prezzi elettrici, viene proposto un approccio basato sulla combinazione di previsioni. Questo metodo, finalizzato a migliorare l’accuratezza previsiva, si è dimostrato utile in molti studi empirici, ma finora non è stato usato nel contesto della previsione dei prezzi elettrici. In questa tesi sono state usate diverse tecniche di combinazione. L’approccio più semplice consiste nel dare lo stesso peso a tutte le previsioni ottenute dai singoli modelli. Altre procedure di combinazione di previsioni sono di tipo adattivo, poichè utilizzano coefficienti non costanti. In questo contesto, sono stati considerati i metodi di Bates & Granger (1969). I modelli usati nella combinazione sono stati scelti, per ciascuna stagione di previsione, con il metodo model confidence set (MCS) descritto in Hansen et al. (2003, 2005) e successivamente ridotti con il metodo forecasts encompassing di Fair & Shiller (1990). Per ciascuna ora considerata, i risultati sottolineano che i modelli si comportano in modo diverso a seconda della stagione di previsione. Questa caratteristica giustifica l’applicazione dei modelli di combinazione di previsioni ad un livello stagionale. In questa tesi vengono presentati risultati promettenti in questa direzione. Considerando le statistiche basate sull’errore di previsione, i risultati delle combinazioni sono stati confrontati con i migliori risultati ottenuti dai singoli modelli in ciascun periodo previsivo. Il vantaggio della procedura proposta deriva dal fatto che combinando le previsioni ad un livello stagionale, si ottengono previsioni di accuratezza superiore o uguale rispetto alle previsioni individuali.

Statistiche Download - Aggiungi a RefWorks
Tipo di EPrint:Tesi di dottorato
Relatore:Bordignon, Silvano
Correlatore:Bunn, Derek W. - Lisi, Francesco
Dottorato (corsi e scuole):Ciclo 21 > Scuole per il 21simo ciclo > SCIENZE STATISTICHE
Data di deposito della tesi:NON SPECIFICATO
Anno di Pubblicazione:31 Luglio 2009
Parole chiave (italiano / inglese):Electricity price forecasting, combination of forecasts, Markov regime switching model, time-varying parameter model, equal predictive accuracy tests
Settori scientifico-disciplinari MIUR:Area 13 - Scienze economiche e statistiche > SECS-S/03 Statistica economica
Struttura di riferimento:Dipartimenti > Dipartimento di Scienze Statistiche
Codice ID:2147
Depositato il:10 Mar 2010 10:54
Simple Metadata
Full Metadata
EndNote Format

Bibliografia

I riferimenti della bibliografia possono essere cercati con Cerca la citazione di AIRE, copiando il titolo dell'articolo (o del libro) e la rivista (se presente) nei campi appositi di "Cerca la Citazione di AIRE".
Le url contenute in alcuni riferimenti sono raggiungibili cliccando sul link alla fine della citazione (Vai!) e tramite Google (Ricerca con Google). Il risultato dipende dalla formattazione della citazione.

Aiolfi, M. & Favero, C. A. (2005). Model uncertainty, thick modeling and thepredictability of stock returns. Journal of Forecasting 24, 233–254. Cerca con Google

Altavilla, C. & De Grauwe, P. (2006). Forecasting and combining compet-ing models of exchange rate determination. CESifo Working Paper Series No.5,CESifo GmbH. Cerca con Google

Arciniegas, A. I. & Arciniegas Rueda, I. E. (2008). Forecasting short-termpower prices in the ontario electricity market (oem) with a fuzzy logic basedinference system. Utilities Policy 16, 39–48. Cerca con Google

Bates, J. M. & Granger, C. W. J. (1969). The combination of forecasts. OR20, 451–468. Cerca con Google

Becker, R. & Clements, A. E. (2008). Are combination forecasts of S&P 500volatility statistically superior? International Journal of Forecasting 24, 122–133. Cerca con Google

Bierbrauer, M., Tr¨uck, S. & Weron, R. (2004). Modeling electricity priceswith regime switching models. Lecture Notes in Computer Science 3039, 859–867. Cerca con Google

Bowden, N. & Payne, J. E. (2008). Short term forecasting of electricity pricesfor miso hubs: Evidence from arima-egarch models. Energy Economics 30, 3186–3197. Cerca con Google

Box, J. E. P. & Jenkins, G. M. (1976). Time Series Analysis, Forecasting andControl. Revised Edition. Holden Day, San Francisco. Cerca con Google

Brown, R. L., Durbin, J. & Evans, J. M. (1975). Techniques for testing theconstancy of regression relationships over time. Journal of the Royal StatisticalSociety. Series B (Methodological) 37, 149–192. Cerca con Google

Bunn, D. W. (1975). A bayesian approach to the linear combination of forecasts.Operational Research Quarterly (1970-1977) 26, 325–329. Cerca con Google

Bunn, D. W. (1977). A comparative evaluation of the outperformance and min-imum variance procedures for the linear synthesis of forecasts. Operational ResearchQuarterly (1970-1977) 28, 653–662. Cerca con Google

Bunn, D. W. (1978). A simplification of the matrix beta distribution for combiningestimators. Journal of the Operational Research Society 29, 1013–1016. Cerca con Google

Bunn, D. W. (1985). Statistical efficiency on the linear combination of forecasts.International Journal of Forecasting 1, 151–163. Cerca con Google

Bunn, D. W. (2000). Forecasting loads and prices in competitive power markets.Proceedings of the IEEE 88, 163–169. Cerca con Google

Bunn, D. W. (2004). Modelling prices in competitive electricity markets. Wiley. Cerca con Google

Bunn, D. W. & Karakatsani, N. (2003). Forecasting electricity prices. WorkingPaper, London Business School. Cerca con Google

Cartea, A. & Figueroa, M. G. (2005). Pricing in electricity markets: A meanreverting jump diffusion model with seasonality. Applied Mathematical Finance12, 313–335. Cerca con Google

Catal˜ao, J. P. S., Mariano, S. J. P. S., Mendes, V. M. S. & Ferreira, L.A. F. M. (2007). Short-term electricity prices forecasting in a competitive market:A neural network approach. Electric Power Systems Research 77, 1297–1304. Cerca con Google

Chen, Z. & Yang, Y. (2007). Time series models for forecasting: Testing orcombining? Studies in Nonlinear Dynamics & Econometrics 11, 1385–1385. Cerca con Google

Clark, T. E. & McCracken, M. W. (2007). Combining forecasts from nestedmodels. Finance and Economics Discussion Series 2007-43, Board of Governorsof the Federal Reserve System (U.S.). Cerca con Google

Clark, T. E. & West, K. D. (2006). Using out-of-sample mean squared predic-tion errors to test the martingale difference hypothesis. Journal of Econometrics135, 155–186. Cerca con Google

Clemen, R. T. (1989). Combining forecasts: A review and annotated bibliography.International Journal of Forecasting 5, 559–583. Cerca con Google

Clemen, R. T. &Winkler, R. L. (1986). Combining economic forecasts. Journalof Business and Economic Statistics 4, 39–46. Cerca con Google

Clemens, M. P. & Hendry, D. F. (1998). Forecasting Economic Time Series.Cambridge University Press. Cerca con Google

Conejo, A. J., Contreras, J., Esp´ınola, R. & Plazas, M. A. (2005a). Fore-casting electricity prices for a day-ahead pool-based electric energy market. InternationalJournal of Forecasting 21, 435–462. Cerca con Google

Conejo, A. J., Plazas, M. A., Esp´ınola, R. & Molina, A. B. (2005b). Day-ahead electricity prices forecasting using the wavelet transform and arima models.IEEE Transactions on Power Systems 20, 1035–1042. Cerca con Google

Contreras, J., Esp´ınola, R., Nogales, F. J. & Conejo, A. J. (2003). Arimamodels to predict next-day electricity prices. IEEE Transactions on Power Systems18, 1014–1020. Cerca con Google

Coutts, J. A., Roberts, J. & Mills, T. C. (1997). Parameter stability in themarket model: tests and time varying parameter estimation with uk data. TheStatistician 46, 57–70. Cerca con Google

Crespo Cuaresma, J., Hlouskova, J., Kossmeier, S. & Obersteiner, M.(2004). Forecasting electricity spot-prices using linear univariate time-series mod-els. Applied Energy 77, 87–106. Cerca con Google

de Jong, C. (2006). The nature of power spikes: A regime-switch approach. Studiesin Nonlinear Dynamics & Econometrics 10, Article 3. Cerca con Google

de Jong, C. & Huisman, R. (2002). Option formulas for mean-reverting powerprices with spikes. Research paper, Erasmus Research Institute of Management(ERIM), Erasmus University Rotterdam. Cerca con Google

de Menezes, L. M., Bunn, D.W. & Taylor, L.W. (2000). Review of guidelinesfor the use of combined forecasts. European Journal of Operational Research 120,190–204. Cerca con Google

Deng, S. (2000). Stochastic models of energy commodity prices and their ap-plications: mean-reversion with jumps and spikes. Power Working Paper 73,University of California Energy Institute. Cerca con Google

Diebold, F. X. & Mariano, R. S. (1995). Comparing predictive accuracy. Journalof Business and Economic Statistics 13, 253–263. Cerca con Google

Duffie, D., Gray, S. & Hoang, P. (1998). Volatility in energy prices. InManaging Energy Price Risk. Risk Publications. Cerca con Google

Dunis, C., Laws, J. & Chauvin, S. (2001). The use of market data and modelcombinations to improve forecast accuracy. In Development in Forecasts Combinationand Portfolio Choice, Dunis, Timmermann & Moody, eds. Wiley, Oxford. Cerca con Google

Durbin, J. & Koopman, S. J. (2001). Time Series Analysis by State SpaceMethods. Oxford University Press. Cerca con Google

Escribano, A., Pe˜na, J. I. & Villaplana, P. (2002). Modeling electric-ity prices: international evidence. Working paper 02-27, University Carlos III,Madrid. Cerca con Google

Fair, R. C. & Shiller, R. J. (1990). Comparing information in forecasts fromeconometric models. American Economic Review 80, 375–389. Cerca con Google

Fezzi, C. (2007a). Econometric models for the analysis of electricity markets. Phdthesis in statistical methodology for scientific research, XIX cycle, Department ofStatistics, University of Bologna. Cerca con Google

Fezzi, C. (2007b). Forecasting electricity prices: the importance of being unsta-ble. 4th International conference: The European Electricity Market. EEM-07.Challenge of the Unification. Cerca con Google

Friedman, J. H. (1984). A variable span scatterplot smoother. Tecnical Report 5,Laboratory for Computentional Statistics, Stanford University. Cerca con Google

Garcia, R. C., Contreras, J., van Akkeren, M. & Garcia, J. B. C. (2005). Cerca con Google

A garch forecasting model to predict day-ahead electricity prices. IEEE Transactionson Power Systems 20, 867–874. Cerca con Google

Geman, H. & Roncoroni, A. (2006). Understanding the fine structure of elec-tricity prices. Journal of Business 79. Cerca con Google

Granger, C. W. J. (2008). Non-linear models: where do we go next - time varyingparameter models? Studies in Nonlinear Dynamics & Econometrics 12, Article1. Cerca con Google

Granger, C. W. J. & Jeon, Y. (2004). Thick modeling. Economic Modelling21, 323–343. Cerca con Google

Guirguis, H. S. & Felder, F. A. (2004). Further advances in forecasting day-ahead electricity prices using time series models. KIEE International Transactionson PE 4-A, 159–166. Cerca con Google

Gupta, S. & Wilton, P. C. (1987). Combination of forecasts: An extension.Management Science 33, 356–372. Cerca con Google

Guthrie, G. & Videbeck, S. (2002). Hight frequency electricity spot price dy-namics: an intra-day approach. Working paper, New Zeland Institute for thestudy of competition and regulation. Cerca con Google

Haldrup, N. & Nielsen, M. O. (2006a). A regime switching long memory modelfor electricity prices. Journal of Econometrics 135, 349–76. Cerca con Google

Hamilton, J. D. (1994). Time Series Analysis. Princeton University Press, Prince-ton. Cerca con Google

Hansen, P. R., Lunde, A. & Nason, L. M. (2003). Choosing the best volatilitymodels: The model confidence set approach. Oxford Bulletin of Economics andStatistics 65, 839–861. Supplement. Cerca con Google

Hansen, P. R., Lunde, A. & Nason, L. M. (2005). Model confidence sets forforecasting models. Working Paper 2005-7, Federal Reserve Bank of Atlanta. Cerca con Google

Harvey, D., Leybourne, S. & Newbold, P. (1997). Testing the equality ofprediction mean squared errors. International Journal of Forecasting 13, 281–291. Cerca con Google

Huisman, R. & Mahieu, R. (2003). Regime jumps in electricity prices. EnergyEconomics 25, 425–34. Cerca con Google

Hyndman, R. J. & Koehler, A. B. (2006). Another look at measures of forecastaccuracy. International Journal of Forecasting 22, 679–688. Cerca con Google

Kanamura, T. & ¯ Ohashi, K. (2008). On transition probabilities of regime switch-ing in electricity prices. Energy Economics 30, 1158–1172. Cerca con Google

Karakatsani, N. & Bunn, D. W. (2004). Modelling stochastic volatility inhigh-frequency spot electricity prices. Working paper, London Business School. Cerca con Google

Karakatsani, N. & Bunn, D. W. (2008a). Intra-day and regime-switching dy-namics in electricity price formation. Energy Economics 30, 1776–1797. Cerca con Google

Karakatsani, N. & Bunn, D. W. (2008b). Forecasting electricity prices: Theimpact of fundamentals and time-varying coefficients. 24, 764–785. InternationalJournal of Forecasting. Cerca con Google

Kim, C. J. (1994). Dynamic linear models with markov-switching. Journal ofEconometrics 60, 1–22. Cerca con Google

Knittel, C. R. & Roberts, M. R. (2005). An empirical examination of restruc-tured electricity prices. Energy Economics 27, 791–817. Cerca con Google

Kosater, P. &Mosler, K. (2006). Can markov regime-switching models improvepower-price forecasts? evidence from german daily power prices. Applied Energy 83, 943–958. Cerca con Google

Kwiatkowski, D., Phillips, P. C. B., Schmidt, P. & Shin, Y. (1992). Testingthe null hypothesis of stationarity against the alternative of a unit root. Journalof Econometrics 54, 159–178. Cerca con Google

Lucia, J. J. & Schwartz, E. S. (2002). Electricity prices and power derivatives:Evidence from the nordic power exchange. Review of Derivatives Research 5,5–50. Cerca con Google

Makridakis, S. &Winkler, R. L. (1983). Averages of forecasts: Some empiricalresults. Management Science 29, 987–996. Cerca con Google

Marcellino, M. (2004). Forecast pooling for short time series of macroeconomicvariables. Oxford Bulletin of Economic and Statistics 6, 91–112. Cerca con Google

Misiorek, A., Trueck, S. & Weron, R. (2006). Point and interval forecastingof spot electricity prices: linear vs. non-linear time series models. Studies inNonlinear Dynamics & Econometrics 10, Article 2. Cerca con Google

Mount, T. D., Ning, Y. & Cai, X. (2006). Predicting price spikes in electricitymarkets using a regime-switching model with time-varying parameters. EnergyEconomics 28, 62–80. Cerca con Google

Newbold, P. & Granger, C. W. J. (1974). Experience with forecasting uni-variate time series and the combination of forecasts (with discussion). Journal ofthe Royal Statistical Society, Series A 137, 131–164. Cerca con Google

Ng, S. & Perron, P. (1995). Unit root tests in ARMA models with data-dependent methods for the selection of the truncation lag. Journal of the AmericanStatistical Association 90, 268–281. Cerca con Google

Nogales, F., Contreras, J., Conejo, A. J., Esp´ınola, R. & Plazas, M. A.(2002). Forecasting next-day electricity prices by time series models. IEEE Transactionson Power Systems 17, 342–348. Cerca con Google

Pedregal, D. J. & Trapero, J. R. (2007). Electricity prices forecasting by auto-matic dynamic harmonic regression models. Energy Conversion and Management48, 1710–1719. Cerca con Google

Phillips, P. C. B. & Perron, P. (1988). Testing for unit roots in time seriesregression. Biometrika 75, 335–346. Cerca con Google

Pilipovic, D. (1998). Energy Risk: Valuing and Managing Energy Derivatives.McGraw-Hill, New York. Cerca con Google

Pino, R., Parreno, J., Gomez, A. & Priore, P. (2008). Forecasting next-day Cerca con Google

price of electricity in the Spanish energy market using artificial neural networks.Engineering Applications of Artificial Intelligence 21, 53–62. Cerca con Google

Ramanathan, R., Engle, R., Granger, C. W. J., Vahid-Araghi, F. &Brace, C. (1997). Short-run forecasting of electricity loads and peaks. InternationalJournal of Forecasting 13, 161–174. Cerca con Google

Reid, D. J. (1968). Combining three estimates of gross domestic product. Economica35, 431–444. Cerca con Google

Reid, D. J. (1969). A comparative study of time series prediction techniques oneconomic data. Ph.D. thesis, University of Nottingham, Nottingham. Cerca con Google

Riedel, S. & Gabrys, B. (2005). Evolving multilevel forecast combination models- an experimental study. In Proceedings of NiSIS’2005 Symposium. Albufeira,Portugal. Cerca con Google

Said, S. E. & Dickey, D. (1984). Testing for unit roots in autoregressive moving-average models with unknown order. Biometrika 71, 599–607. Cerca con Google

S´anchez, I. (2008). Adaptive combination of forecasts with application to wind energy. International Journal of Forecasting 24, 679–693. Cerca con Google

Serati, M., Manera, M. & Plotegher, M. (2007). Modelling electricity prices:from the state of the art to a draft of a new proposal. LIUC Papers in Economics210, Cattaneo University (LIUC). Cerca con Google

Serletis, A. & Shahmoradi, A. (2006). Measuring and testing natural gasand electricity markets volatility: Evidence from alberta’s deregulated markets.Studies in Nonlinear Dynamics & Econometrics 10, Article 10. Cerca con Google

Smith, J. & Wallis, K. F. (2009). A simple explanation of the forecast combi-nation puzzle. Oxford Bulletin of Economics and Statistics 71, 331–355. Cerca con Google

Stock, J. H. & Watson, M. W. (2001). A comparison of linear and nonlinearunivariate models for forecasting macroeconomic time series. In Festschrift inHonour of Clive Granger, R. F. Engle & H. White, eds. Cambridge UniversityPress, Cambridge, pp. 1–44. Cerca con Google

Stock, J. H. & Watson, M. W. (2004). Combination forecasts of output growthin a seven-country data set. Journal of Forecasting 23, 405–430. Cerca con Google

Swider, D. J. & Weber, C. (2007). Extended ARMA models for estimatingprice developments on day-ahead electricity markets. Electric Power Research77, 583–593. Cerca con Google

Timmermann, A. G. (2006). Forecast combinations. In Handbook of EconomicForecasting, G. C. W. J. Elliot, G & A. Timmermann, eds., vol. 1. Amsterdam:Elsevier, pp. 135–196. Cerca con Google

Vahidinasab, V., Jadid, S. & Kazemi, A. (2008). Day-ahead price forecastingin restructured power systems using artificial neural networks. Electric PowerSystems Research 78, 1332–1342. Cerca con Google

Weron, R. (2002). Measuring long-range dependence in electricity prices. EmpiricalScience of Financial Fluctuations , 110–119. Cerca con Google

Weron, R. (2005). Heavy tails and electricity prices. Invited paper presented at theDeutsche Bundesbank’s 2005 Annual Fall Conference (Eltville, 10-12 November2005). Cerca con Google

Weron, R. (2006). Modelling and Forecasting Electricity Loads and Prices: AStatistical Approach. Wiley, Chichester. Cerca con Google

Weron, R., Bierbrauer, M. & Tr¨uck, S. (2004a). Modeling electricity prices:jump diffusion and regime switching. Physica A 336, 39–48. Cerca con Google

Weron, R. & Misiorek, A. (2005). Forecasting spot electricity prices with timeseries models. In Proceedings Volume. International Conference ”The EuropeanElectricity Market EEM-05”, May 10-12, 2005, Lodz, Poland. Cerca con Google

Weron, R. & Misiorek, A. (2006). Short-term electricity price forecasting withtime series models: A review and evaluation. In Complex Electricity Markets,W. Mielczarski, ed., The European Power Industry Supply, chap. 10. Lodz, pp.231–254. Cerca con Google

Weron, R. & Przyby lowicz, B. (2000). Hurst analysis of electricity price dy-namics. Physica A 283, 462–468. Cerca con Google

Weron, R., Simonsen, I. & Wilman, P. (2004b). Modeling highly volatile andseasonal markets: evidence from the nord pool electricity market. Applications ofEconophysics , 182–191. Cerca con Google

West, K. D. (1996). Asymptotic inference about predictive ability. Econometrica64, 1067–1084. Cerca con Google

Winkler, R. L. & Makridakis, S. (1983). The combination of forecasts. Journalof the Royal Statistical Society Series A 146, 150–157. Cerca con Google

Yang, Y. (2004). Combining forecasting procedures: some theoretical results.Econometric Theory 20, 176–222. Cerca con Google

Download statistics

Solo per lo Staff dell Archivio: Modifica questo record