nep-for New Economics Papers
on Forecasting
Issue of 2014‒03‒15
sixteen papers chosen by
Rob J Hyndman
Monash University

  1. Merging quantile regression with forecast averaging to obtain more accurate interval forecasts of Nord Pool spot prices By Jakub Nowotarski; Rafal Weron
  2. Predicting Exchange Rates Out of Sample: Can Economic Fundamentals Beat the Random Walk? By Jiahan Li; Ilias Tsiakas; Wei Wang
  3. A review of electricity price forecasting: The past, the present and the future By Rafal Weron
  4. The uncertainty of conditional returns, volatilities and correlations in DCC models By Diego Fresoli; Esther Ruiz
  5. Forecasting the oil-gasoline price relationship: should we care about the Rockets and the Feathers? By Andrea Bastianin; Marzio Galeotti; Matteo Manera
  6. Forecasting recessions in real time By Knut Are Aastveit; Anne Sofie Jore; Francesco Ravazzolo
  7. Generalised Density Forecast Combinations By N. Fawcett; G. Kapetanios; J. Mitchell; S. Price
  8. Exchange Rate Predictability in a Changing World By Joseph P. Byrne; Dimitris Korobilis; Pinho J. Ribeiro
  9. Forecasting the intraday market price of money By Andrea Monticini; Francesco Ravazzolo
  11. Model Averaging in Predictive Regressions By Liu, Chu-An; Kuo, Biing-Shen
  12. Data-based priors for vector autoregressions with drifting coefficients By Dimitris Korobilis
  13. Risk Assessment of the Brazilian FX Rate By Wagner Piazza Gaglianone; Jaqueline Terra Moura Marins
  14. Modelling spatiotemporal variability of temperature By Xiaofeng Cao; Ostap Okhrin; Martin Odening; Matthias Ritter
  15. Netconomics: Novel Forecasting Techniques from the Combination of Big Data, Network Science and Economics By Andreas Joseph; Irena Vodenska; Eugene Stanley; Guanrong Chen
  16. Do Google Trend data contain more predictability than price returns? By Damien Challet; Ahmed Bel Hadj Ayed

  1. By: Jakub Nowotarski; Rafal Weron
    Abstract: We evaluate a recently proposed method for constructing prediction intervals, which utilizes the concept of quantile regression (QR) and a pool of point forecasts of different time series models.We find that in terms of interval forecasting of Nord Pool day-ahead prices the new QR-based approach significantly outperforms prediction intervals obtained from standard, as well as, semi-parametric autoregressive time series models.
    Keywords: Electricity spot price; Prediction interval; Quantile regression; Forecasts combination
    JEL: C22 C24 C53 Q47
    Date: 2014–03–10
  2. By: Jiahan Li (University of Notre Dame, USA); Ilias Tsiakas (University of Guelph, Canada); Wei Wang (Fifth Third Bank, USA)
    Abstract: This paper shows that economic fundamentals can generate reliable out-of-sample forecasts for exchange rates when prediction is based on a "kitchen-sink" regression that incorporates multiple predictors. The key to establishing predictability is estimating the kitchen-sink regression with the elastic-net shrinkage method, which improves performance by reducing the effect of less informative predictors in out-of-sample forecasting. Using statistical and economic measures of predictability, we show that our approach outperforms alternative models, including the random walk, individual exchange rate models, a kitchen-sink regression estimated with ordinary least squares, standard forecast combinations and popular ad-hoc strategies such as momentum and the 1/N strategy.
    Keywords: Exchange Rates; Out-of-Sample Forecasting; Elastic Net; Combined Forecasts
    JEL: F31 F37 G11 G15 G17
    Date: 2014–02
  3. By: Rafal Weron
    Abstract: A variety of methods and ideas have been tried for electricity price forecasting (EPF), 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.
    Keywords: Electricity price forecasting, Day-ahead market, Seasonality, Autoregression, Neural network, Factor model, Forecasts combination
    JEL: C22 C24 C38 C53 Q47
    Date: 2014–03–10
  4. By: Diego Fresoli; Esther Ruiz
    Abstract: When forecasting conditional correlations that evolve according to a Dynamic Conditional Correlation (DCC) model, only point forecasts can be obtained at each moment of time. In this paper, we analyze the finite sample properties of a bootstrap procedure to approximate the density of these forecasts that also allows obtaining conditional densities for future returns and volatilities. The procedure is illustrated by obtaining conditional forecast intervals and regions of returns, volatilities andcorrelations in the context of a system of daily exchange rates returns of the Euro, Japanese Yen and Australian Dollar against the US Dollar
    Keywords: Bootstrap forecast intervals, Forecast regions, Dynamic Conditional Correlation, Exchange rates, Realized correlation, Resampling methods
    Date: 2014–02
  5. By: Andrea Bastianin; Marzio Galeotti; Matteo Manera
    Abstract: According to the Rockets and Feathers hypothesis (RFH), the transmission mechanism of positive and negative changes in the price of crude oil to the price of gasoline is asymmetric. Although there have been many contributions documenting that downstream prices are more reactive to increases than to decreases in upstream prices, little is known about the forecasting performance of econometric models incorporating asymmetric price transmission from crude oil to gasoline. In this paper we fill this gap by comparing point, sign and probability forecasts from a variety of Asymmetric-ECM (A-ECM) and Threshold Autoregressive ECM (TAR-ECM) specifications against a standard ECM. Forecasts from A-ECM and TAR-ECM subsume the RFH, while the ECM implies symmetric price transmission from crude oil to gasoline. We quantify the forecast accuracy gains due to incorporating the RFH in predictive models for the prices of gasoline and diesel. We show that the RFH is useless for point forecasting, while it can be exploited to produce more accurate sign and probability forecasts. Finally, we highlight that the forecasting performance of the estimated models is time-varying.
    Keywords: Asymmetries, Forecast Evaluation, Gasoline, Crude Oil, Rockets and Feathers
    JEL: C22 C32 C53 Q40 Q47
    Date: 2014–03
  6. By: Knut Are Aastveit (Norges Bank (Central Bank of Norway)); Anne Sofie Jore (Norges Bank (Central Bank of Norway)); Francesco Ravazzolo (Norges Bank (Central Bank of Norway))
    Abstract: We review several methods to define and forecast classical business cycle turning points in Norway. In the paper we compare the Bry - Boschan rule (BB) with a Markov Switching model (MS), using alternative vintages of Norwegian Gross Domestic Product (GDP) as the business cycle indicator. The timing of business cycles depends on the vintage and the method used. BB provides the most reasonable definition of business cycles. The forecasting exercise, where the models are augmented with surveys or financial indicators, respectively, leads to the conclusion that the BB rule applied to density forecasts of GDP augmented with either the consumer confidence index or a financial conditions index provides the most timely predictions of peaks. For troughs, augmenting with surveys or financial indicators does not increase forecastability.
    Keywords: Forecast densities, Turning points, Real-time data
    JEL: C32 C52 C53 E37 E52
    Date: 2014–02–13
  7. By: N. Fawcett; G. Kapetanios; J. Mitchell; S. Price
    Abstract: Density forecast combinations are becoming increasingly popular as a means of improving forecast `accuracy’, as measured by a scoring rule. In this paper we generalise this literature by letting the combination weights follow more general schemes. Sieve estimation is used to optimise the score of the generalised density combination where the combination weights depend on the variable one is trying to forecast. Specific attention is paid to the use of piecewise linear weight functions that let the weights vary by region of the density. We analyse these schemes theoretically, in Monte Carlo experiments and in an empirical study. Our results show that the generalised combinations outperform their linear counterparts.
    Keywords: Density Forecasting, Model Combination, Scoring Rules
    JEL: C53
    Date: 2014–03
  8. By: Joseph P. Byrne (Department of Economics, Heriot-Watt University, UK); Dimitris Korobilis (Department of Economics, Adam Smith Business School, University of Glasgow, UK); Pinho J. Ribeiro (Department of Economics, Adam Smith Business School, University of Glasgow, UK)
    Abstract: An expanding literature articulates the view that Taylor rules are helpful in predicting exchange rates. In a changing world however, Taylor rule parameters may be subject to structural instabilities, for example during the Global Financial Crisis. This paper forecasts exchange rates using such Taylor rules with Time Varying arameters (TVP) estimated by Bayesian methods. In core out-of-sample results, we improve upon a random walk benchmark for at least half, and for as many as eight out of ten, of the currencies considered. This contrasts with a constant parameter Taylor rule model that yields a more limited improvement upon the benchmark. In further results, Purchasing Power Parity and Uncovered Interest Rate Parity TVP models beat a random walk benchmark, implying our methods have some generality in exchange rate prediction.
    Keywords: Exchange Rate Forecasting; Taylor Rules; Time-Varying Parameters; Bayesian Methods
    JEL: C53 E52 F31 F37 G17
    Date: 2014–02
  9. By: Andrea Monticini (Dipartimento di Economia e Finanza, Università Cattolica del Sacro Cuore); Francesco Ravazzolo (Norges Bank and BI Norwegian Business School)
    Abstract: Central banks' operations and eciency arguments would suggest that the intraday interest rate should be set to zero. However, a liquidity crisis introduces frictions related to news, which can cause an upward jump of the intraday rate. This paper documents that these dynamics can be partially predicted during turbulent times. Long memory approaches or a combination of them to account for model uncertainty outperform random walk, autoregressive and moving average benchmarks in terms of point and density forecasting. The relative accuracy is higher when the full distribution is predicted. We also document that such statistical accuracy can provide economic gains in investment strategies based on lending in the intraday market.
    Keywords: interbank market, intraday interest rate, forecasting, density forecasting, policy tools.
    JEL: C22 C53 E4 E5
    Date: 2014–02
    Date: 2014
  11. By: Liu, Chu-An; Kuo, Biing-Shen
    Abstract: This paper considers forecast combination in a predictive regression. We construct the point forecast by combining predictions from all possible linear regression models given a set of potentially relevant predictors. We propose a frequentist model averaging criterion, an asymptotically unbiased estimator of the mean squared forecast error (MSFE), to select forecast weights. In contrast to the existing literature, we derive the MSFE in a local asymptotic framework without the i.i.d. normal assumption. This result allows us to decompose the MSFE into the bias and variance components and also to account for the correlations between candidate models. Monte Carlo simulations show that our averaging estimator has much lower MSFE than alternative methods such as weighted AIC, weighted BIC, Mallows model averaging, and jackknife model averaging. We apply the proposed method to stock return predictions.
    Keywords: Forecast combination, Local asymptotic theory, Plug-in estimators.
    JEL: C52 C53
    Date: 2014–03–07
  12. By: Dimitris Korobilis
    Abstract: This paper proposes full-Bayes priors for time-varying parameter vector autoregressions (TVP-VARs) which are more robust and objective than existing choices proposed in the literature. We formulate the priors in a way that they allow for straightforward posterior computation, they require minimal input by the user, and they result in shrinkage posterior representations, thus, making them appropriate for models of large dimensions. A comprehensive forecasting exercise involving TVP-VARs of different dimensions establishes the usefulness of the proposed approach.
    Keywords: TVP-VAR, shrinkage, data-based prior, forecasting
    JEL: C11 C22 C32 C52 C53 C63 E17 E58
    Date: 2014–01
  13. By: Wagner Piazza Gaglianone; Jaqueline Terra Moura Marins
    Abstract: In this paper, we construct several multi-step-ahead density forecasts for the foreign exchange (FX) rate based on statistical, financial data and economic-driven approaches. The objective is to go beyond the standard conditional mean investigation of the FX rate and (for instance) allow for asymmetric responses of covariates (e.g. financial data or economic fundamentals) in respect to exchange rate movements. We also provide a toolkit to evaluate out-of-sample density forecasts and select models for risk analysis purposes. An empirical exercise for the Brazilian FX rate is provided. Overall, the results suggest that no single model properly accounts for the entire density in all considered forecast horizons. Nonetheless, the GARCH model as well as the option-implied approach seem to be more suitable for short-run purposes (until three months), whereas the survey-based and some economic-driven models appear to be more adequate for longer horizons (such as one year)
    Date: 2014–01
  14. By: Xiaofeng Cao; Ostap Okhrin; Martin Odening; Matthias Ritter
    Abstract: Forecasting temperature in time and space is an important precondition for both the design of weather derivatives and the assessment of the hedging effectiveness of index based weather insur-ance. In this article, we show how this task can be accomplished by means of Kriging techniques. Moreover, we compare Kriging with a dynamic semiparametric factor model (DSFM) that has been recently developed for the analysis of high dimensional financial data. We apply both methods to comprehensive temperature data covering a large area of China and assess their performance in terms of predicting a temperature index at an unobserved location. The results show that the DSFM performs worse than standard Kriging techniques. Moreover, we show how geographic basis risk inherent to weather derivatives can be mitigated by regional diversification.
    Keywords: weather insurance, semiparametric model, factor model, Kriging, geographic basis risk
    JEL: C14 C53 G32
    Date: 2014–02
  15. By: Andreas Joseph; Irena Vodenska; Eugene Stanley; Guanrong Chen
    Abstract: The combination of the network theoretic approach with recently available abundant economic data leads to the development of novel analytic and computational tools for modelling and forecasting key economic indicators. The main idea is to introduce a topological component into the analysis, taking into account consistently all higher-order interactions. We present three basic methodologies to demonstrate different approaches to harness the resulting network gain. First, a multiple linear regression optimisation algorithm is used to generate a relational network between individual components of national balance of payment accounts. This model describes annual statistics with a high accuracy and delivers good forecasts for the majority of indicators. Second, an early-warning mechanism for global financial crises is presented, which combines network measures with standard economic indicators. From the analysis of the cross-border portfolio investment network of long-term debt securities, the proliferation of a wide range of over-the-counter-traded financial derivative products, such as credit default swaps, can be described in terms of gross-market values and notional outstanding amounts, which are associated with increased levels of market interdependence and systemic risk. Third, considering the flow-network of goods traded between G-20 economies, network statistics provide better proxies for key economic measures than conventional indicators. For example, it is shown that a country's gate-keeping potential, as a measure for local power, projects its annual change of GDP generally far better than the volume of its imports or exports.
    Date: 2014–03
  16. By: Damien Challet; Ahmed Bel Hadj Ayed
    Abstract: Using non-linear machine learning methods and a proper backtest procedure, we critically examine the claim that Google Trends can predict future price returns. We first review the many potential biases that may influence backtests with this kind of data positively, the choice of keywords being by far the greatest culprit. We then argue that the real question is whether such data contain more predictability than price returns themselves: our backtest yields a performance of about 17bps per week which only weakly depends on the kind of data on which predictors are based, i.e. either past price returns or Google Trends data, or both.
    Date: 2014–03

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.