nep-for New Economics Papers
on Forecasting
Issue of 2019‒09‒09
fifteen papers chosen by
Rob J Hyndman
Monash University

  1. Forecasting Public Investment Using Daily Stock Returns By Morita, Hiroshi
  2. QCNN: Quantile Convolutional Neural Network By G\'abor Petneh\'azi
  3. The finer points of model comparison in machine learning: forecasting based on russian banks’ data By Denis Shibitov; Mariam Mamedli
  4. Are Bitcoins price predictable? Evidence from machine learning techniques using technical indicators By Samuel Asante Gyamerah
  5. Predicting systemic financial crises with recurrent neural networks By Tölö, Eero
  6. Forecasting crude oil prices with DSGE models By Michał Rubaszek
  7. HATS: A Hierarchical Graph Attention Network for Stock Movement Prediction By Raehyun Kim; Chan Ho So; Minbyul Jeong; Sanghoon Lee; Jinkyu Kim; Jaewoo Kang
  8. Predicting Returns With Text Data By Zheng Tracy Ke; Bryan T. Kelly; Dacheng Xiu
  9. Micro Forecasting By Mat\'u\v{s} Maciak; Ostap Okhrin; Michal Pe\v{s}ta
  10. Forecasting CBOT Corn Futures Price with Dynamic Model Averaging: The Roles of Fundamentals, Financial Markets, and Economics Environment By Xiong, Tao
  11. Agricultural Loan Delinquency Prediction Using Machine Learning Methods By Chen, Jian; Katchova, Ani
  12. Predict Food Security with Machine Learning: Application in Eastern Africa By Zhou, Yujun; Baylis, Kathy
  13. Medical and nursery care with endogenous health and longevity By Schünemann, Johannes; Strulik, Holger; Trimborn, Timo
  14. Predicting Consumer Default: A Deep Learning Approach By Stefania Albanesi; Domonkos F. Vamossy
  15. Predicting financial distress of companies: Comparison between multivariate discriminant analysis and multilayer perceptron for Tunisian case By Fayçal Mraihi; Inane Kanzari

  1. By: Morita, Hiroshi
    Abstract: This paper investigates the predictability of public investment in Japan using the daily excess stock returns of the construction industry, to contribute to the recent discussion on fiscal foresight. To examine the relationship between monthly public investment and daily stock returns without any prior time aggregation, we employ the VAR model with MIDAS regression and estimate the optimal weights for connecting high-frequency and low-frequency data in addition to VAR coefficients and the variance-covariance structure. We find that the VAR model with MIDAS regression reduces the mean square prediction error in out-of-sample forecasting by approximately 15% and 2.5% compared to the no-change forecast and VAR model forecasting with prior time aggregation, respectively. Moreover, using the local projection method, we find evidence of the fiscal news shock estimated in our proposed model delaying positive effects on output, consumption, hours worked, and real wage when news shocks actually result in increasing public investment. This finding suggests the New Keynesian structure of the Japanese economy.
    Keywords: MIDAS regression, fiscal foresight, stock returns, local projection method
    JEL: C22 C53 E62
    Date: 2019–08
  2. By: G\'abor Petneh\'azi
    Abstract: A dilated causal one-dimensional convolutional neural network architecture is proposed for quantile regression. The model can forecast any arbitrary quantile, and it can be trained jointly on multiple similar time series. An application to Value at Risk forecasting shows that QCNN outperforms linear quantile regression and constant quantile estimates.
    Date: 2019–08
  3. By: Denis Shibitov (Bank of Russia, Russian Federation); Mariam Mamedli (Bank of Russia, Russian Federation)
    Abstract: We evaluate the forecasting ability of machine learning models to predict bank license withdrawal and the violation of statutory capital and liquidity requirements (capital adequacy ratio N1.0, common equity Tier 1 adequacy ratio N1.1, Tier 1 capital adequacy ratio N1.2, N2 instant and N3 current liquidity). On the basis of 35 series from the accounting reports of Russian banks, we form two data sets of 69 and 721 variables and use them to build random forest and gradient boosting models along with neural networks and a stacking model for different forecasting horizons (1, 2, 3, 6, 9 months). Based on the data from February 2014 to October 2018 we show that these models with fine-tuned architectures can successfully compete with logistic regression usually applied for this task. Stacking and random forest generally have the best forecasting performance comparing to the other models. We evaluate models with commonly used performance metrics (ROC-AUC and F1) and show that, depending on the task, F1-score could be better at defining the model’s performance. Comparison of the results depending on the metrics applied and types of cross-validation used illustrate the importance of choosing the appropriate metric for performance evaluation and the cross-validation procedure, which accounts for the characteristics of the data set and the task under consideration. The developed approach shows the advantages of non-linear methods for bank regulation tasks and provides the guidelines for the application of machine learning algorithms to these tasks.
    Keywords: machine learning, random forest, neural networks, gradient boosting, forecasting, bank supervision
    JEL: C53 C52 C5
    Date: 2019–08
  4. By: Samuel Asante Gyamerah
    Abstract: The uncertainties in future Bitcoin price make it difficult to accurately predict the price of Bitcoin. Accurately predicting the price for Bitcoin is therefore important for decision-making process of investors and market players in the cryptocurrency market. Using historical data from 01/01/2012 to 16/08/2019, machine learning techniques (Generalized linear model via penalized maximum likelihood, random forest, support vector regression with linear kernel, and stacking ensemble) were used to forecast the price of Bitcoin. The prediction models employed key and high dimensional technical indicators as the predictors. The performance of these techniques were evaluated using mean absolute percentage error (MAPE), root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R-squared). The performance metrics revealed that the stacking ensemble model with two base learner (random forest and generalized linear model via penalized maximum likelihood) and support vector regression with linear kernel as meta-learner was the optimal model for forecasting Bitcoin price. The MAPE, RMSE, MAE, and R-squared values for the stacking ensemble model were 0.0191%, 15.5331 USD, 124.5508 USD, and 0.9967 respectively. These values show a high degree of reliability in predicting the price of Bitcoin using the stacking ensemble model. Accurately predicting the future price of Bitcoin will yield significant returns for investors and market players in the cryptocurrency market.
    Date: 2019–09
  5. By: Tölö, Eero
    Abstract: We consider predicting systemic financial crises one to five years ahead using recurrent neural networks. The prediction performance is evaluated with the Jorda-Schularick-Taylor dataset, which includes the crisis dates and relevant macroeconomic series of 17 countries over the period 1870-2016. Previous literature has found simple neural network architectures to be useful in predicting systemic financial crises. We show that such predictions can be greatly improved by making use of recurrent neural network architectures, especially suited for dealing with time series input. The results remain robust after extensive sensitivity analysis.
    JEL: G21 C45 C52
    Date: 2019–08–27
  6. By: Michał Rubaszek (SGH Warsaw School of Economics)
    Date: 2019–08–16
  7. By: Raehyun Kim; Chan Ho So; Minbyul Jeong; Sanghoon Lee; Jinkyu Kim; Jaewoo Kang
    Abstract: Many researchers both in academia and industry have long been interested in the stock market. Numerous approaches were developed to accurately predict future trends in stock prices. Recently, there has been a growing interest in utilizing graph-structured data in computer science research communities. Methods that use relational data for stock market prediction have been recently proposed, but they are still in their infancy. First, the quality of collected information from different types of relations can vary considerably. No existing work has focused on the effect of using different types of relations on stock market prediction or finding an effective way to selectively aggregate information on different relation types. Furthermore, existing works have focused on only individual stock prediction which is similar to the node classification task. To address this, we propose a hierarchical attention network for stock prediction (HATS) which uses relational data for stock market prediction. Our HATS method selectively aggregates information on different relation types and adds the information to the representations of each company. Specifically, node representations are initialized with features extracted from a feature extraction module. HATS is used as a relational modeling module with initialized node representations. Then, node representations with the added information are fed into a task-specific layer. Our method is used for predicting not only individual stock prices but also market index movements, which is similar to the graph classification task. The experimental results show that performance can change depending on the relational data used. HATS which can automatically select information outperformed all the existing methods.
    Date: 2019–08
  8. By: Zheng Tracy Ke; Bryan T. Kelly; Dacheng Xiu
    Abstract: We introduce a new text-mining methodology that extracts sentiment information from news articles to predict asset returns. Unlike more common sentiment scores used for stock return prediction (e.g., those sold by commercial vendors or built with dictionary-based methods), our supervised learning framework constructs a sentiment score that is specifically adapted to the problem of return prediction. Our method proceeds in three steps: 1) isolating a list of sentiment terms via predictive screening, 2) assigning sentiment weights to these words via topic modeling, and 3) aggregating terms into an article-level sentiment score via penalized likelihood. We derive theoretical guarantees on the accuracy of estimates from our model with minimal assumptions. In our empirical analysis, we text-mine one of the most actively monitored streams of news articles in the financial system—the Dow Jones Newswires—and show that our supervised sentiment model excels at extracting return-predictive signals in this context.
    JEL: C53 C58 G10 G11 G12 G14 G17
    Date: 2019–08
  9. By: Mat\'u\v{s} Maciak; Ostap Okhrin; Michal Pe\v{s}ta
    Abstract: Forecasting costs is now a front burner in empirical economics. We propose an unconventional tool for stochastic prediction of future expenses based on the individual (micro) developments of recorded events. Consider a firm, enterprise, institution, or state, which possesses knowledge about particular historical events. For each event, there is a series of several related subevents: payments or losses spread over time, which all leads to an infinitely stochastic process at the end. Nevertheless, the issue is that some already occurred events do not have to be necessarily reported. The aim lies in forecasting future subevent flows coming from already reported, occurred but not reported, and yet not occurred events. Our methodology is illustrated on quantitative risk assessment, however, it can be applied to other areas such as startups, epidemics, war damages, advertising and commercials, digital payments, or drug prescription as manifested in the paper. As a theoretical contribution, inference for infinitely stochastic processes is developed. In particular, a non-homogeneous Poisson process with non-homogeneous Poisson processes as marks is used, which includes for instance the Cox process as a special case.
    Date: 2019–08
  10. By: Xiong, Tao
    Keywords: Demand and Price Analysis
    Date: 2019–06–25
  11. By: Chen, Jian; Katchova, Ani
    Keywords: Agricultural Finance
    Date: 2019–06–25
  12. By: Zhou, Yujun; Baylis, Kathy
    Keywords: International Development
    Date: 2019–06–25
  13. By: Schünemann, Johannes (Faculty of Economics and Social Sciences); Strulik, Holger; Trimborn, Timo
    Abstract: For the population over 65, nursery care expenditures constitute on average the largest share in total health expenditures. In this paper, we distinguish between medical care, intended to improve one›s state of health, and personal care required for daily routine. Personal care can be either carried out autonomously or by a third party. In the course of aging, autonomous personal care is eventually substituted by nursery care. We set up a life-cycle model in which individuals are subject to physiological aging, calibrate it with data from gerontology, and analyze the interplay between medical and nursery care. We replicate health behavior and life expectancy of individuals and in particular the empirically observed patterns of medical and nursery care expenditure. We then analyze the impact of better health and rising life expectancy, triggered by rising income and medical progress, on the expected cost of nursery care in the future. We predict an elasticity of nursery care expenditure with respect to life expectancy of 1/3. In terms of present value at age 20, life-time nursery care expenditure is predicted to decline with rising life expectancy.
    JEL: D11 D91 I12 J11
    Date: 2019–09–03
  14. By: Stefania Albanesi; Domonkos F. Vamossy
    Abstract: We develop a model to predict consumer default based on deep learning. We show that the model consistently outperforms standard credit scoring models, even though it uses the same data. Our model is interpretable and is able to provide a score to a larger class of borrowers relative to standard credit scoring models while accurately tracking variations in systemic risk. We argue that these properties can provide valuable insights for the design of policies targeted at reducing consumer default and alleviating its burden on borrowers and lenders, as well as macroprudential regulation.
    Date: 2019–08
  15. By: Fayçal Mraihi (Higher School of Economics and Business Sciences of Tunis); Inane Kanzari (Higher School of Economics and Business Sciences of Tunis)
    Abstract: In this study, we try to develop a model for predicting corporate default based on a multivariate discriminant analysis (ADM) and a multilayer perceptron (MLP). The two models are applied to the Tunisian cases. Our sample consists of 212 companies in the various industries (106 ‘healthy’ companies and 106 “distressed” companies) over the period 2005-2010. The results of the use of a battery of 87 ratios showed that 16 ratios can build the model and that liquidity and solvency have more weight than profitability and management in predicting the distress. Despite the slight superiority of the results provided by the MLP model, on the control sample, the results provided by the two models are good either in terms of correct percentage of classification or in terms of stability of discriminating power over time and space.
    Date: 2019–08–21

This nep-for issue is ©2019 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.