nep-for New Economics Papers
on Forecasting
Issue of 2014‒01‒10
nine papers chosen by
Rob J Hyndman
Monash University

  1. Computing electricity spot price prediction intervals using quantile regression and forecast averaging By Jakub Nowotarski; Rafal Weron
  2. Bayesian Inference and Forecasting in the Stationary Bilinear Model By Roberto Leon-Gonzalez; Fuyu Yang
  3. Forecasting of daily electricity prices with factor models: Utilizing intra-day and inter-zone relationships By Katarzyna Maciejowska; Rafal Weron
  4. The structure of a machine-built forecasting system By Jiaqi Chen; Michael L. Tindall
  5. Forecasting Bank Credit Ratings By Periklis Gogas; Theophilos Papadimitriou; Anna Agrapetidou
  6. A 14-Variable Mixed-Frequency VAR Model By Beauchemin, Kenneth
  7. The Baltic Dry Index: Cyclicalities, Forecasting and Hedging Strategies By Fotis Papailias; Dimitrios D. Thomakos
  8. Forecasting with Factor Models: A Bayesian Model Averaging Perspective By Dimitris, Korobilis
  9. General equilibrium dynamics with naive and sophisticated hyperbolic consumers in an overlapping generations economy By Takeshi Ojima

  1. By: Jakub Nowotarski; Rafal Weron
    Abstract: We examine possible accuracy gains from forecast averaging in the context of interval forecasts of electricity spot prices. First, we test whether constructing empirical prediction intervals (PI) from combined electricity spot price forecasts leads to better forecasts than those obtained from individual methods. Next, we propose a new method for constructing PI, which utilizes the concept of quantile regression (QR) and a pool of point forecasts of individual (i.e. not combined) time series models. While the empirical PI from combined forecasts do not provide significant gains, the QR based PI are found to be more accurate than those of the best individual model - the smoothed nonparametric autoregressive model.
    Keywords: Prediction interval; Quantile regression; Forecasts combination; Electricity spot price
    JEL: C22 C24 C53 Q47
    Date: 2013–12–31
  2. By: Roberto Leon-Gonzalez (National Graduate Institute for Policy Studies); Fuyu Yang (University of East Anglia)
    Abstract: A stationary bilinear (SB) model can be used to describe processes with a time-varying degree of persistence that depends on past shocks. The SB model has been used to model highly persistent but stationary macroeconomic time series such as inflation. This study develops methods for Bayesian inference, model comparison, and forecasting in the SB model. Using U.K. inflation data, we find that the SB model outperforms the random walk and first order autoregressive AR(1) models, in terms of root mean squared forecast errors for the one-step-ahead out-of-sample forecast. In addition, the SB model is superior to these two models in terms of predictive likelihood for almost all of the forecast observations.
    Date: 2014–01
  3. By: Katarzyna Maciejowska; Rafal Weron
    Abstract: We show that incorporating the intra-day and inter-zone relationships of electricity prices in the Pennsylvania--New Jersey--Maryland (PJM) Interconnection improves the accuracy of short- and medium-term forecasts of average daily prices for a major PJM market hub -- the Dominion Hub in Virginia, U.S. The forecasting performance of four multivariate models calibrated to hourly and/or zonal day-ahead prices is evaluated and compared with that of a univariate model, which uses only average daily data for the Dominion Hub. The multivariate competitors include a restricted vector autoregressive model and three factor models with the common and idiosyncratic components estimated using principal components in a semiparametric setup. The results indicate that there are forecast improvements from incorporating the additional information, essentially for all considered forecast horizons ranging from one day to two months, but only when the correlation structure of prices across locations and hours is modeled using factor models.
    Keywords: Wholesale electricity price; Forecasting; Vector autoregression; Factor model; Principal components; PJM market
    JEL: C32 C38 C53 Q47
    Date: 2013–12–30
  4. By: Jiaqi Chen; Michael L. Tindall
    Abstract: This paper describes the structure of a rule-based econometric forecasting system designed to produce multi-equation econometric models. The paper describes the functioning of a working system which builds the econometric forecasting equation for each series submitted and produces forecasts of the series. The system employs information criteria and cross validation in the equation building process, and it uses Bayesian model averaging to combine forecasts of individual series. The system outperforms standard benchmarks for a variety of national economic datasets.
    Keywords: Econometrics
    Date: 2013
  5. By: Periklis Gogas (Department of Economics, Democritus University of Thrace, Greece); Theophilos Papadimitriou (Department of Economics, Democritus University of Thrace, Greece); Anna Agrapetidou (Department of Economics, Democritus University of Thrace, Greece)
    Abstract: Purpose-This study presents an empirical model designed to forecast bank credit ratings. For this reason we use the long term ratings provided by Fitch in 2012. Our sample consists of 92 U.S. banks and publicly available information from their financial statements from 2008 to 2011. Methodology -First, in the effort to select the most informative regressors from a long list of financial variables and ratios we use stepwise least squares and select several alternative sets of variables. Then these sets of variables are used in an ordered probit regression setting to forecast the long term credit ratings. Findings-Under this scheme, the forecasting accuracy of our best model reaches 83.70% when 9 explanatory variables are used. Originality/value- The results indicate that bank credit ratings largely rely on historical data making them respond sluggishly and after any financial problems were already known to the public.
    Date: 2013–11
  6. By: Beauchemin, Kenneth (Federal Reserve Bank of Minneapolis)
    Abstract: This paper describes recent modifications to the mixed-frequency model vector autoregression (MF-VAR) constructed by Schorfheide and Song (2012). The changes to the model are restricted solely to the set of variables included in the model; all other aspects of the model remain unchanged. Forecast evaluations are conducted to gauge the accuracy of the revised model to standard benchmarks and the original model.
    Keywords: Bayesian Vector Autoregression; Forecasting
    JEL: C11 C32 C53
    Date: 2013–12–19
  7. By: Fotis Papailias (Queen's University Management School, UK); Dimitrios D. Thomakos (Department of Economics, University of Peloponnese, Greece; Rimini Centre for Economic Analysis, Rimini, Italy)
    Abstract: The cyclical properties of the annual growth of the Baltic Dry Index (BDI) and their implications for short-to-medium term forecasting performance are investigated. We show that the BDI has a cyclical pattern which has been stable except for a period after the 2007 crisis. This pattern has implications for improved forecasting and strategic management on the future path of the BDI. To illustrate the practicality of our results, we perform an investment exercise that depends on the predicted signs. The empirical evidence supports the presence of the cyclical component and the ability of using forecast signs for improved risk management.
    Keywords: Baltic Dry Index, Commodities, Concordance, Cyclical Analysis, Forecasting, Hedging, Turning Points
    Date: 2013–12
  8. By: Dimitris, Korobilis
    Abstract: We use Bayesian factor regression models to construct a financial conditions index (FCI) for the U.S. Within this context we develop Bayesian model averaging methods that allow the data to select which variables should be included in the FCI or not. We also examine the importance of different sources of instability in the factors, such as stochastic volatility and structural breaks. Our results indicate that ignoring structural breaks in the loadings can be quite costly in terms of the forecasting performance of the FCI. Additionally, Bayesian model averaging can improve in specific cases the performance of the FCI, by means of discarding irrelevant financial variables during the estimation of the factor.
    Keywords: financial stress; stochastic search variable selection; early-warning system; forecasting
    JEL: C11 C22 C52 C53 C63 E17 G01 G17
    Date: 2013–01
  9. By: Takeshi Ojima
    Abstract: Using an overlapping generations model, this paper describes interactions between naive and sophisticated hyperbolic discounters in general equilibrium. The naifs, who overestimate their future propensity to save and hence over-forecast the future equilibrium asset prices, are exploited through capital transactions by sophisticates, who correctly forecast the future asset prices by incorporating the naifsf mis-forecasts. Due to the capital losses, the naifs fall into bankruptcy when they are highly present-biased, highly patient, and having a low population density. Under generous conditions, the equilibrium is shown to be globally stable and Pareto inefficient in the ex-post sense.
    Date: 2013–09

This nep-for issue is ©2014 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.