nep-for New Economics Papers
on Forecasting
Issue of 2020‒08‒17
twelve papers chosen by
Rob J Hyndman
Monash University

  1. Tail risk forecasting using Bayesian realized EGARCH models By Vica Tendenan; Richard Gerlach; Chao Wang
  2. One model or many? Exchange rates determinants and their predictive capabilities. By Piotr Dybka
  3. Forecast Performance in Times of Terrorism By Jonathan Benchimol; Makram El-Shagi
  4. Reliable real-time output gap estimates based on a modified Hamilton filter By Quast, Josefine; Wolters, Maik H.
  5. Predicting bond return predictability By Daniel Borup; Jonas N. Eriksen; Mads M. Kjær; Martin Thyrsgaard
  6. Too similar to combine? On negative weights in forecast combination By Radchenko, Peter; Vasnev, Andrey; Wang, Wendun
  7. Higher Moment Constraints for Predictive Density Combinations By Pauwels, Laurent; Radchenko, Peter; Vasnev, Andrey
  8. Modelling GDP for Sudan using ARIMA By Moahmed Hassan, Hisham; Haleeb, Amin
  9. The time-varying risk of Italian GDP By Fabio Busetti; Michele Caivano; Davide Delle Monache; Claudia Pacella
  10. The hard problem of prediction for conflict prevention By Hannes Mueller; Christopher Rauh
  11. Identification of Volatility Proxies as Expectations of Squared Financial Return By Sucarrat, Genaro
  12. Selective Attention in Exchange Rate Forecasting By Svatopluk Kapounek; Zuzana Kucerova; Evzen Kocenda

  1. By: Vica Tendenan; Richard Gerlach; Chao Wang
    Abstract: This paper develops a Bayesian framework for the realized exponential generalized autoregressive conditional heteroskedasticity (realized EGARCH) model, which can incorporate multiple realized volatility measures for the modelling of a return series. The realized EGARCH model is extended by adopting a standardized Student-t and a standardized skewed Student-t distribution for the return equation. Different types of realized measures, such as sub-sampled realized variance, sub-sampled realized range, and realized kernel, are considered in the paper. The Bayesian Markov chain Monte Carlo (MCMC) estimation employs the robust adaptive Metropolis algorithm (RAM) in the burn in period and the standard random walk Metropolis in the sample period. The Bayesian estimators show more favourable results than maximum likelihood estimators in a simulation study. We test the proposed models with several indices to forecast one-step-ahead Value at Risk (VaR) and Expected Shortfall (ES) over a period of 1000 days. Rigorous tail risk forecast evaluations show that the realized EGARCH models employing the standardized skewed Student-t distribution and incorporating sub-sampled realized range are favored, compared to a range of models.
    Date: 2020–08
  2. By: Piotr Dybka
    Abstract: In this paper the Dynamic Bayesian Model Averaging (DMA) algorithm is used to establish the key determinants of the nominal exchange rates of 5 currencies: CAD, EUR, GBP, CHF and JPY against the US dollar. My results indicate that the importance of the variables in the exchange rate forecasting can substantially differ in time. Even among the set of developed countries, there are visible differences in the set of key determinants of the exchange rate. However, the lagged value of the exchange rate remains always an important variable indicating significant persistence in the exchange rate time series. Furthermore, the PPP rate, Terms of Trade (TOT) and output per worker are also variables that have high Posterior Inclusion Probabilities among the analyzed countries. My results show that macroeconomic fundamentals are not leading indicators of the exchange rates. As a result, to outperform the random walk (naive) forecast of the exchange rate using the macroeconomic fundamentals, a good quality of the forecast of the explanatory variables is required.
    Keywords: Exchange rates, forecasting, Bayesian Model Averaging
    JEL: C11 C33 F14 F15
    Date: 2020–07
  3. By: Jonathan Benchimol; Makram El-Shagi
    Abstract: Governments, central banks and private companies make extensive use of expert and market-based forecasts in their decision-making processes. These forecasts can be affected by terrorism, a factor that should be considered by decision-makers. We focus on terrorism as a mostly endogenously driven form of political uncertainty and assess the forecasting performance of market-based and professional inflation and exchange rate forecasts in Israel. We show that expert forecasts are better than market-based forecasts, particularly during periods of terrorism. However, the performance of both market-based and expert forecasts is significantly worse during such periods. Thus, policymakers should be particularly attentive to terrorism when considering inflation and exchange rate forecasts.
    Keywords: inflation; exchange rate; forecast performance; terrorism; market forecast; expert forecast
    JEL: C53 E37 F37 F51
    Date: 2020–06–26
  4. By: Quast, Josefine; Wolters, Maik H.
    Abstract: We propose a simple modification of Hamilton's (2018) time series filter that yields reliable and economically meaningful real-time output gap estimates. The original filter relies on 8 quarter ahead forecast errors of a simple autoregression of real GDP. While this approach yields a cyclical component that is hardly revised with new incoming data due to the one-sided filtering approach, it does not cover typical business cycle frequencies evenly, but mutes short and amplifies medium length cycles. Further, as the estimated trend contains high frequency noise, it can hardly be interpreted as potential GDP. A simple modification based on the mean of 4 to 12 quarter ahead forecast errors shares the favorable real-time properties of the Hamilton filter, but leads to a much better coverage of typical business cycle frequencies and a smooth estimated trend. Based on output growth and inflation forecasts and a comparison to revised output gap estimates from policy institutions, we find that real-time output gaps based on the modified and the original Hamilton filter are economically much more meaningful measures of the business cycle than those based on other simple statistical trend-cycle decomposition techniques, such as the HP or bandpass filter, and should thus be used preferably.
    Keywords: business cycle measurement,potential output,trend-cycle decomposition,real-time data,inflation forecasting,output growth forecasting
    JEL: C18 E32 E37
    Date: 2020
  5. By: Daniel Borup (Aarhus University, CREATES and the Danish Finance Institute (DFI)); Jonas N. Eriksen (Aarhus University, CREATES and the Danish Finance Institute (DFI)); Mads M. Kjær (Aarhus University and CREATES); Martin Thyrsgaard (Northwestern University and CREATES)
    Abstract: We document predictable shifts in bond return predictability. Bond returns are predictable in high (low) economic activity (uncertainty) states, implying that the expectations hypothesis of the term structure holds periodically. These predictable performance differences, established using a new multivariate test for equal conditional predictive ability, can be used in real-time to improve out-of-sample bond risk premia estimates and investors’ economic value by means of a novel dynamic forecast combination scheme. Consistent with standard financial theory, the resulting forecasts are strongly countercyclical and peaks in recessions. The empirical findings are explained within a non-linear term structure model.
    Keywords: Bond excess returns, forecasting, state-dependencies, multivariate test, equal conditional predictive ability
    JEL: C12 C52 E43 E44 G12
    Date: 2020–08–04
  6. By: Radchenko, Peter; Vasnev, Andrey; Wang, Wendun
    Abstract: This paper provides the first thorough investigation of the negative weights that can emerge when combining forecasts. The usual practice in the literature is to ignore or trim negative weights, i.e., set them to zero. This default strategy has its merits, but it is not optimal. We study the problem from a variety of different angles, and the main conclusion is that negative weights emerge when highly correlated forecasts with similar variances are combined. In this situation, the estimated weights have large variances, and trimming reduces the variance of the weights and improves the combined forecast. The threshold of zero is arbitrary and can be improved. We propose an optimal trimming threshold, i.e., an additional tuning parameter to improve forecasting performance. The effects of optimal trimming are demonstrated in simulations. In the empirical example using the European Central Bank Survey of Professional Forecasters, we find that the new strategy performs exceptionally well and can deliver improvements of more than 10% for inflation, up to 20% for GDP growth, and more than 20% for unemployment forecasts relative to the equal-weight benchmark.
    Keywords: Forecast combination; Optimal weights; Negative weight; Trimming
    Date: 2020–07–28
  7. By: Pauwels, Laurent; Radchenko, Peter; Vasnev, Andrey
    Abstract: The majority of financial data exhibit asymmetry and heavy tails, which makes forecasting the entire density critically important. Recently, a forecast combination methodology has been developed to combine predictive densities. We show that combining individual predictive densities that are skewed and/or heavy-tailed results in significantly reduced skewness and kurtosis. We propose a solution to over- come this problem by deriving optimal log score weights under Higher-order Moment Constraints (HMC). The statistical properties of these weights are investigated theoretically and through a simulation study. Consistency and asymptotic distribution results for the optimal log score weights with and without high moment constraints are derived. An empirical application that uses the S&P 500 daily index returns illustrates that the proposed HMC weight density combinations perform very well relative to other combination methods.
    Keywords: Forecast combinations; Predictive densities; Moment constraints; Financial data
    JEL: C53 C58
    Date: 2020–05–01
  8. By: Moahmed Hassan, Hisham; Haleeb, Amin
    Abstract: This paper aims to obtain an appropriate ARIMA model for the Sudan GDP using the Box- Jenkins methodology during the period 1960-2018 the various ARIMA models with different order of autoregressive and moving-average terms were compared. The appropriate model for Sudan is an ARIMA (1,1,1), the results of an in-sample forecast showed that the relative and predicted values were within the range of 5%, and the forecasting effectiveness of this model, its relatively adequate and efficient in modeling the annual GDP of the Sudan.
    Keywords: ARIMA Modelling, Box-Jenkins methodology, forecasting, GDP, Sudan.
    JEL: E00 E01 E60
    Date: 2020
  9. By: Fabio Busetti (Bank of Italy); Michele Caivano (Bank of Italy); Davide Delle Monache (Bank of Italy); Claudia Pacella (Bank of Italy)
    Abstract: The uncertainty surrounding economic forecasts is generally related to multiple sources of risks, of domestic and foreign origin. This paper studies the predictive distribution of Italian GDP growth as a function of selected risk indicators, related to both financial and real economic developments. The conditional distribution is characterized by means of expectile regressions. Expectiles are closely related to the Expected Shortfall, a well-known measure of risk with desirable properties. Here a decomposition of Expected Shortfall in terms of contributions of different indicators is proposed, which allows to track over time the main drivers of risk. Our analysis of the predictive distribution of GDP confirms that financial conditions are relevant for the left tail of the distribution but it also highlights that indicators of global trade and uncertainty have strong explanatory power for both left and right tail. Their usefulness is supported also in a pseudo real-time predictive context. Overall, our findings suggest that Italian GDP risks have been mostly driven by foreign developments around the Great Recession, by domestic financial conditions at the time of the Sovereign Debt Crisis and by economic policy uncertainty in more recent years.
    Keywords: asymmetric least squares, expectiles, density forecasts, GDP growth, risks
    JEL: C53 E17
    Date: 2020–07
  10. By: Hannes Mueller (Institut d'Analisi Economica (CSIC)); Christopher Rauh (Université de Montréal)
    Abstract: There is a rising interest in conflict prevention and this interest provides a strong motivation for better conflict forecasting. A key problem of conflict forecasting for prevention is that predicting the start of conflict in previously peaceful countries is extremely hard. To make progress in this hard problem this project exploits both supervised and unsupervised machine learning. Specifically, the latent Dirichlet allocation (LDA) model is used for feature extraction from 3.8 million newspaper articles and these features are then used in a random forest model to predict conflict. We find that several features are negatively associated with the outbreak of conflict and these gain importance when predicting hard onsets. This is because the decision tree uses the text features in lower nodes where they are evaluated conditionally on conflict history, which allows the random forest to adapt to the hard problem and provides useful forecasts for prevention.
    Date: 2019–04
  11. By: Sucarrat, Genaro
    Abstract: Volatility proxies like Realised Volatility (RV) are extensively used to assess the forecasts of squared financial return produced by Autoregressive Conditional Heteroscedasticity (ARCH) models. But are volatility proxies identified as expectations of the squared return? If not, then the results of these comparisons can be misleading, even if the proxy is unbiased. Here, a tripartite distinction between strong, semi-strong and weak identification of a volatility proxy as an expectation of squared return is introduced. The definition implies that semi-strong and weak identification can be studied and corrected for via a multiplicative transformation. Well-known tests can be used to check for identification and bias, and Monte Carlo simulations show they are well-sized and powerful -- even in fairly small samples. As an illustration, twelve volatility proxies used in three seminal studies are revisited. Half of the proxies do not satisfy either semi-strong or weak identification, but their corrected transformations do. Correcting for identification does not always reduce the bias of the proxy, so there is a tradeoff between the choice of correction and the resulting bias.
    Keywords: GARCH models, financial time-series econometrics, volatility forecasting, Realised Volatility
    JEL: C18 C22 C53 C58
    Date: 2020–07–20
  12. By: Svatopluk Kapounek (Mendel University in Brno, Faculty of Business and Economics); Zuzana Kucerova (Mendel University in Brno, Faculty of Business and Economics); Evzen Kocenda (Institute of Economic Studies, Charles University)
    Abstract: We analyze the exchange rate forecasting performance under the assumption of selective attention. Although currency markets react to a variety of different information, we hypothesize that market participants process only a limited amount of information. Our analysis includes more than 100,000 news articles relevant to the six most-traded foreign exchange currency pairs for the period of 1979-2016. We employ a dynamic model averaging approach to reduce model selection uncertainty and to identify time-varying probability to include regressors in our models. Our results show that considering selective attention improves forecasting results. Specifically, we document a growing impact of foreign trade and monetary policy news on the Euro/United States of America dollar currency pair following the global financial crisis. Overall, our results point to the existence of selective attention in the case of most currency pairs.
    Keywords: exchange rate, selective attention, news, dynamic model averaging
    JEL: F33 C11
    Date: 2020–07

This nep-for issue is ©2020 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.