nep-for New Economics Papers
on Forecasting
Issue of 2022‒01‒10
eleven papers chosen by
Rob J Hyndman
Monash University

  1. Using Polls to Forecast Popular Vote Share for US Presidential Elections 2016 and 2020: An Optimal Forecast Combination Based on Ensemble Empirical Model By Easaw, Joshy; Fang, Yongmei; Heravi, Saeed
  2. CBO’s Economic Forecasting Record: 2021 Update By Congressional Budget Office
  3. Generative Adversarial Network (GAN) and Enhanced Root Mean Square Error (ERMSE): Deep Learning for Stock Price Movement Prediction By Ashish Kumar; Abeer Alsadoon; P. W. C. Prasad; Salma Abdullah; Tarik A. Rashid; Duong Thu Hang Pham; Tran Quoc Vinh Nguyen
  4. Modeling and forecasting international tourism demand in Zimbabwe: a bright future for Zimbabwe's tourism industry By NYONI, THABANI
  5. Development of an Ensemble of Models for Predicting Socio-Economic Indicators of the Russian Federation using IRT-Theory and Bagging Methods By Kitova, Olga; Savinova, Victoria
  6. Safe Havens, Machine Learning, and the Sources of Geopolitical Risk: A Forecasting Analysis Using Over a Century of Data By Rangan Gupta; Sayar Karmakar; Christian Pierdzioch
  7. What should be taken into consideration when forecasting oil implied volatility index? By Delis, Panagiotis; Degiannakis, Stavros; Giannopoulos, Kostantinos
  8. Expert Aggregation for Financial Forecasting By Bri\`ere Marie; Alasseur Cl\'emence; Joseph Mikael; Carl Remlinger
  9. Inefficiency in Social Security Trust Funds Forecasts By Kajal Lahiri; Junyan Zhang; Yongchen Zhao
  10. Forex Trading Volatility Prediction using NeuralNetwork Models By Shujian Liao; Jian Chen; Hao Ni
  11. Forecasting Social Unrest: A Machine Learning Approach By Sandile Hlatshwayo; Chris Redl

  1. By: Easaw, Joshy (Cardiff Business School); Fang, Yongmei (College of Mathematics and Informatics, South China Agricultural University, China); Heravi, Saeed (Cardiff Business School)
    Abstract: This study introduces the Ensemble Empirical Mode Decomposition (EEMD) technique to forecasting popular vote share. The technique is useful when using polling data, which is pertinent when none of the main candidates is the incumbent. Our main interest in this study is the short- and long-term forecasting and, thus, we consider from the short forecast horizon of 1-day to three months ahead. The EEMD technique is used to decompose the election data for the two most recent US presidential elections; 2016 and 2020 US. Three models, Support Vector Machine (SVM), Neural Network (NN) and ARIMA models are then used to predict the decomposition components. The final hybrid model is then constructed by comparing the prediction performance of the decomposition components. The predicting performance of the combination model are compared with the benchmark individual models, SVM, NN, and ARIMA. In addition, this compared to the single prediction market IOWA Electronic Markets. The results indicated that the prediction performance of EEMD combined model is better than that of individual models.
    Keywords: Forecasting Popular Votes Shares; Electoral Poll; Forecast combination, Hybrid model; Support Vector Machine
    Date: 2021–12
  2. By: Congressional Budget Office
    Abstract: CBO evaluates its two-year and five-year economic forecasts from as early as 1976 and compares them with analogous forecasts from the Administration and the Blue Chip consensus—an average of about 50 private-sector forecasts.
    JEL: C53 H20
    Date: 2021–12–09
  3. By: Ashish Kumar; Abeer Alsadoon; P. W. C. Prasad; Salma Abdullah; Tarik A. Rashid; Duong Thu Hang Pham; Tran Quoc Vinh Nguyen
    Abstract: The prediction of stock price movement direction is significant in financial circles and academic. Stock price contains complex, incomplete, and fuzzy information which makes it an extremely difficult task to predict its development trend. Predicting and analysing financial data is a nonlinear, time-dependent problem. With rapid development in machine learning and deep learning, this task can be performed more effectively by a purposely designed network. This paper aims to improve prediction accuracy and minimizing forecasting error loss through deep learning architecture by using Generative Adversarial Networks. It was proposed a generic model consisting of Phase-space Reconstruction (PSR) method for reconstructing price series and Generative Adversarial Network (GAN) which is a combination of two neural networks which are Long Short-Term Memory (LSTM) as Generative model and Convolutional Neural Network (CNN) as Discriminative model for adversarial training to forecast the stock market. LSTM will generate new instances based on historical basic indicators information and then CNN will estimate whether the data is predicted by LSTM or is real. It was found that the Generative Adversarial Network (GAN) has performed well on the enhanced root mean square error to LSTM, as it was 4.35% more accurate in predicting the direction and reduced processing time and RMSE by 78 secs and 0.029, respectively. This study provides a better result in the accuracy of the stock index. It seems that the proposed system concentrates on minimizing the root mean square error and processing time and improving the direction prediction accuracy, and provides a better result in the accuracy of the stock index.
    Date: 2021–11
    Abstract: This paper, which is the first of its kind in the case of Zimbabwe, uses annual time series on international tourism demand in Zimbabwe from 1980 to 2019, to model and forecast the demand for international tourism using the Box – Jenkins ARIMA approach. This research has been guided by the following objectives: to analyze international tourism trends in Zimbabwe over the study period, to develop and estimate a reliable international tourism forecasting model for Zimbabwe based on the Box-Jenkins ARIMA technique and to project international tourism demand in Zimbabwe over the next decade (2020 – 2030). Based on the Akaike Information Criterion (AIC), the study presents the ARIMA (2, 1, 0) model as the optimal model. The ARIMA (2, 1, 0) model proves beyond any reasonable doubt that over the period 2020 to 2030, international tourism demand in Zimbabwe will increase and that indeed, the future of Zimbabwe’s tourism industry is bright. Amongst other policy recommendations, the study advocates for the continued implementation and enforcement of COVID-19 preventive and control measures as well as unwavering support for tourism sector development through policies such as the National Tourism Recovery and Growth Strategy.
    Keywords: Forecasting; international tourism, Zimbabwe
    JEL: L83 Z0
    Date: 2021–12–02
  5. By: Kitova, Olga; Savinova, Victoria
    Abstract: This article describes the application of the bagging method to build a forecast model for the socio-economic indicators of the Russian Federation. This task is one of the priorities within the framework of the Federal Project "Strategic Planning", which implies the creation of a unified decision support system capable of predicting socio-economic indicators. This paper considers the relevance of the development of forecasting models, examines and analyzes the work of researchers on this topic. The authors carried out computational experiments for 40 indicators of the socio-economic sphere of the Russian Federation. For each indicator, a linear multiple regression equation was constructed. For the constructed equations, verification was carried out and indicators with the worst accuracy and quality of the forecast were selected. For these indicators, neural network modeling was carried out. Multilayer perceptrons were chosen as the architecture of neural networks. Next, an analysis of the accuracy and quality of neural network models was carried out. Indicators that could not be predicted with a sufficient level of accuracy were selected for the bagging procedure. Bagging was used for weighted averaging of prediction results for neural networks of various configurations. Item Response Theory (IRT) elements were used to determine the weights of the models.
    Keywords: Socio-economic Indicators of the Russian Federation, Forecasting, Bagging, Multiple Linear Regression, Neural Networks, Item Response Theory
    JEL: C45
    Date: 2021–11–25
  6. By: Rangan Gupta (Department of Economics, University of Pretoria, Private Bag X20, Hatfield 0028, South Africa); Sayar Karmakar (Department of Statistics, University of Florida, 230 Newell Drive, Gainesville, FL, 32601, USA); Christian Pierdzioch (Department of Economics, Helmut Schmidt University, Holstenhofweg 85, P.O.B. 700822, 22008 Hamburg, Germany)
    Abstract: We use monthly data covering a century-long sample period (1915-2021) to study whether geopolitical risk helps to forecast subsequent gold returns and gold volatility. We account not only for geopolitical threats and acts, but also for 39 country-specific sources of geopolitical risk. The response of subsequent returns and volatility is heterogeneous across countries and nonlinear. We find that accounting for geopolitical risk at the country level improves forecast accuracy especially when we use random forests to estimate our forecasting models. As an extension, we report empirical evidence on the predictive value of the country-level sources of geopolitical risk for two other candidate safe-haven assets, oil and silver, over the sample periods 1900–2021 and 1915–2021, respectively. Our results have important implications for the portfolio decisions of investors who seek a safe haven in times of heightened geopolitical tensions.
    Keywords: Gold, Geopolitical Risk, Forecasting, Returns, Volatility, Random Forests
    JEL: C22 D80 H56 Q02
    Date: 2022–01
  7. By: Delis, Panagiotis; Degiannakis, Stavros; Giannopoulos, Kostantinos
    Abstract: Crude oil is considered a key commodity in all the economies around the world. This study forecasts the oil volatility index (OVX), which is the market’s expectation of future oil volatility, by incorporating information from other asset classes. The literature does not extensively test the long memory of the targeted volatility. Thus, we estimate the Hurst exponent implementing a rolling window rescaled analysis. We provide evidence for a strong long memory in the implied volatility (IV) indices which justifies the use of the HAR model in obtaining multiple days ahead OVX forecasts. We also define a dynamic model averaging (DMA) structure in the HAR model in order to allow for IV indices from other asset classes to be applicable at different time periods. The implementation of the DMA-HAR models informs forecasters to focus on the major stock market IV indices, and more specifically on the DJIA Volatility Index. Our results lead us to the conclusion that accurate OVX forecasts are obtained for short- and mid-run forecasting horizons. The evaluation framework is not limited to statistical loss functions but also embodies an options straddle trading strategy.
    Keywords: crude oil, implied volatility, HAR modelling, trading strategies, dynamic model averaging, long memory
    JEL: C58 G17 Q47
    Date: 2021–11–26
  8. By: Bri\`ere Marie; Alasseur Cl\'emence; Joseph Mikael; Carl Remlinger
    Abstract: Machine learning algorithms dedicated to financial time series forecasting have gained a lot of interest over the last few years. One difficulty lies in the choice between several algorithms, as their estimation accuracy may be unstable through time. In this paper, we propose to apply an online aggregation-based forecasting model combining several machine learning techniques to build a portfolio which dynamically adapts itself to market conditions. We apply this aggregation technique to the construction of a long-short-portfolio of individual stocks ranked on their financial characteristics and we demonstrate how aggregation outperforms single algorithms both in terms of performances and of stability.
    Date: 2021–11
  9. By: Kajal Lahiri; Junyan Zhang; Yongchen Zhao
    Abstract: We examine forecast accuracy and efficiency of the Social Security Administration’s projections for cost rate, trust fund balance, trust fund ratio made during 1980-2020 with horizons up to 95 years. We find that the reported deterioration in the accuracy of the forecasts during 2010’s has reversed. The level of informational inefficiency was pervasive during 1990-2009, although it shows signs of improvement after 2010.
    Keywords: social security trust funds, long-range solvency forecasts, Nordhaus test, forecast efficiency, fixed-event forecast revisions
    JEL: C53 E37 E66 H55 H68
    Date: 2021
  10. By: Shujian Liao; Jian Chen; Hao Ni
    Abstract: In this paper, we investigate the problem of predicting the future volatility of Forex currency pairs using the deep learning techniques. We show step-by-step how to construct the deep-learning network by the guidance of the empirical patterns of the intra-day volatility. The numerical results show that the multiscale Long Short-Term Memory (LSTM) model with the input of multi-currency pairs consistently achieves the state-of-the-art accuracy compared with both the conventional baselines, i.e. autoregressive and GARCH model, and the other deep learning models.
    Date: 2021–12
  11. By: Sandile Hlatshwayo; Chris Redl
    Abstract: We produce a social unrest risk index for 125 countries covering a period of 1996 to 2020. The risk of social unrest is based on the probability of unrest in the following year derived from a machine learning model drawing on over 340 indicators covering a wide range of macro-financial, socioeconomic, development and political variables. The prediction model correctly forecasts unrest in the following year approximately two-thirds of the time. Shapley values indicate that the key drivers of the predictions include high levels of unrest, food price inflation and mobile phone penetration, which accord with previous findings in the literature.
    Keywords: Social unrest, machine learning.
    Date: 2021–11–05

This nep-for issue is ©2022 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.