nep-for New Economics Papers
on Forecasting
Issue of 2018‒09‒03
eight papers chosen by
Rob J Hyndman
Monash University

  1. Probabilistic electricity price forecasting with NARX networks: Combine point or probabilistic forecasts? By Grzegorz Marcjasz; Bartosz Uniejewski; Rafal Weron
  2. A note on averaging day-ahead electricity price forecasts across calibration windows By Katarzyna Hubicka; Grzegorz Marcjasz; Rafal Weron
  3. Density Forecasts in Panel Models: A semiparametric Bayesian Perspective* By Laura Liu
  4. Forecasting with Dynamic Panel Data Models By Laura Liu; Hyungsik Moon; Frank Schorfheide
  5. Point and Density Forecasts of Oil Returns: The Role of Geopolitical Risks By Vasilios Plakandaras; Rangan Gupta; Wing-Keung Wong
  6. The quanto theory of exchange rates By Kremens, Lukas; Martin, Ian
  7. Efficient forecasting of electricity spot prices with expert and LASSO models By Bartosz Uniejewski; Rafal Weron
  8. Information Content of DSGE Forecasts By Ray Fair

  1. By: Grzegorz Marcjasz; Bartosz Uniejewski; Rafal Weron
    Abstract: A recent electricity price forecasting (EPF) study has shown that the Seasonal Component Artificial Neural Network (SCANN) modeling framework, which consists of decomposing a series of spot prices into a trend-seasonal and a stochastic component, modeling them independently and then combining their forecasts, can yield more accurate point predictions than an approach in which the same non-linear autoregressive NARX-type neural network is calibrated to the prices themselves. Here, considering two novel extensions of the SCANN concept to probabilistic forecasting, we find that (i) efficiently calibrated NARX networks can outperform their autoregressive counterparts, even without combining forecasts from many runs, and that (ii) in terms of accuracy it is better to construct probabilistic forecasts directly from point predictions, however, if speed is a critical issue, running quantile regression on combined point forecasts (i.e., committee machines) may be an option worth considering. Moreover, we confirm an earlier observation that averaging probabilities outperforms averaging quantiles when combining predictive distributions in EPF.
    Keywords: Electricity spot price; Probabilistic forecast; Combining forecasts; Long-term seasonal component; NARX neural network; Quantile regression
    JEL: C14 C22 C45 C51 C53 Q47
    Date: 2018–07–13
  2. By: Katarzyna Hubicka; Grzegorz Marcjasz; Rafal Weron
    Abstract: We propose a novel concept in energy forecasting and show that averaging day-ahead electricity price forecasts of a predictive model across 28- to 728-day calibration windows yields better results than selecting only one 'optimal' window length. Even more significant accuracy gains can be achieved by averaging over a few, carefully selected windows.
    Keywords: Electricity price forecasting; Combining forecasts; Calibration window; Autoregression; NARX neural network; Committee machine; Diebold-Mariano test
    JEL: C14 C22 C45 C51 C53 Q47
    Date: 2018–07–07
  3. By: Laura Liu (Federal Reserve Bank)
    Abstract: This paper constructs individual-specific density forecasts for a panel of firms or households using a dynamic linear model with common and heterogeneous coeficients and cross-sectional heteroskedasticity. The panel considered in this paper features large cross-sectional dimension (N) but short time series (T). Due to short T, traditional methods have difficulty in disentanglingthe heterogeneous parameters from the shocks, which contaminates the estimates of the heterogeneous parameters. To tackle this problem, I assume that there is an underlying distribution of heterogeneous parameters, model this distribution nonparametrically allowing for correlation between heterogeneous parameters and initial conditions as well as individual-specific regressors, and then estimate this distribution by pooling the information from the whole cross-section together. I develop a simulation-based posterior sampling algorithm specifically addressing the nonparametric density estimation of unobserved heterogeneous parameters. I prove that both the estimated common parameters and the estimated distribution of the heterogeneous parameters achieve posterior consistency, and that the density forecasts asymptotically converge to the oracle forecast, an (infeasible) benchmark that is defined as the individual-specific posterior predictive distribution under the assumption that the common parameters and the distribution of the heterogeneous parameters are known. Monte Carlo simulations demonstrate improvements in density forecasts relative to alternative approaches. An application to young firm dynamics also shows that the proposed predictor provides more accurate density predictions.
    Keywords: Bayesian, Semiparametric Methods, Panel Data, Density Forecasts, Posterior Consistency, Young Firms Dynamics
    JEL: C11 C14 C23 C53 L25
    Date: 2017–04–28
  4. By: Laura Liu (Federal Reserve Bank); Hyungsik Moon (Department of Economics, USC); Frank Schorfheide (Department of Economics, University of Pennsylvania)
    Abstract: This paper considers the problem of forecasting a collection of short time series using cross sectional information in panel data. We construct point predictors using Tweedie's formula for the posterior mean of heterogeneous coefficients under a correlated random effects distribution. This formula utilizes cross-sectional information to transform the unit-specific (quasi) maximum likelihood estimator into an approximation of the posterior mean under a prior distribution that equals the population distribution of the random coefficients. We show that the risk of a predictor based on a non-parametric estimate of the Tweedie correction is asymptotically equivalent to the risk of a predictor that treats the correlated-random-effects distribution as known (ratio-optimality). Our empirical Bayes predictor performs well compared to various competitors in a Monte Carlo study. In an empirical application we use the predictor to forecast revenues for a large panel of bank holding companies and compare forecasts that condition on actual and severely adverse macroeconomic conditions.
    Keywords: Bank Stress Tests, Empirical Bayes, Forecasting, Panel Data, Ratio Optimality, Tweedies Formula
    JEL: C11 C14 C23 C53 G21
    Date: 2016–12–21
  5. By: Vasilios Plakandaras (Department of Economics, Democritus University of Thrace, Greece); Rangan Gupta (Department of Economics, University of Pretoria, Pretoria, South Africa); Wing-Keung Wong (Department of Finance, Fintech Center, and Big Data Research Center, Asia University; Department of Medical Research, China Medical University Hospital; Department of Economics and Finance, Hang Seng Management College; Department of Economics, Lingnan University)
    Abstract: We examine the dynamic relationship between oil prices and news-based indices of global geopolitical risks (GPRs), as well as a composite measure of the same for emerging economies, which we develop using Dynamic Model Averaging (DMA). In doing so, we train a number of linear and nonlinear probabilistic models to capture the ability of GPRs in forecasting oil returns. Our empirical findings show that global GPRs associated with wars is the most accurate in forecasting oil returns in the short-run, while composite GPRs emanating from the emerging markets, forecasts oil returns relatively better at medium- to longer-horizons. However, differences across the linear and nonlinear models incorporating information of GPRs are not necessarily markedly different. Given an observe negative relationship between GPRs and oil returns, density forecasts show that increases in GPRs from their initial lower levels, which would imply higher conditional oil returns initially, can predict the resulting increases in oil returns thereafter more accurately compared to the lower end of the conditional distribution, which in turn, corresponds to higher initial levels of GPRs.
    Keywords: Bayesian VAR, Geopolitical Risks, Oil Prices, Dynamic Model Averaging
    JEL: C22 C32 Q41 Q47
    Date: 2018–07
  6. By: Kremens, Lukas; Martin, Ian
    Abstract: We present a new identity that relates expected exchange rate appreciation to a risk-neutral covariance term, and use it to motivate a currency forecasting variable based on the prices of quanto index contracts. We show via panel regressions that the quanto forecast variable is an economically and statistically significant predictor of currency appreciation and of excess returns on currency trades. Out of sample, the quanto variable outperforms predictions based on uncovered interest parity, on purchasing power parity, and on a random walk as a forecaster of differential (dollar-neutral) currency appreciation.
    JEL: F31 F37 F47 G12 G15
    Date: 2018–08–11
  7. By: Bartosz Uniejewski; Rafal Weron
    Abstract: Recent electricity price forecasting (EPF) studies suggest that the least absolute shrinkage and selection operator (LASSO) leads to well performing models, generally better than obtained from other variable selection schemes. Conducting an empirical study involving three expert models, two multi-parameter regression (called baseline) models and four variance stabilizing transformations, we discuss the optimal way of implementing the LASSO. We show that using a complex baseline model and a well chosen variance stabilizing transformation indeed leads to significant accuracy gains compared to the typically used EPF models.
    Keywords: Electricity spot price; Day-ahead market; Long-term seasonal component; LASSO; Automated variable selection; Variance stabilizing transformation
    JEL: C14 C22 C51 C53 Q47
    Date: 2018–06–29
  8. By: Ray Fair
    Abstract: This paper examines the question whether information is contained in forecasts from DSGE models beyond that contained in lagged values, which are extensively used in the models. Four sets of forecasts are examined. The results are encouraging for DSGE forecasts of real GDP. The results suggest that there is information in the DSGE forecasts not contained in forecasts based only on lagged values and that there is no information in the lagged-value forecasts not contained in the DSGE forecasts. The opposite is true for forecasts of the GDP deflator. Keywords: DSGE forecasts, Lagged values JEL Classification Codes: E10, E17, C53
    Date: 2018–08

This nep-for issue is ©2018 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.