nep-for New Economics Papers
on Forecasting
Issue of 2021‒04‒19
eight papers chosen by
Rob J Hyndman
Monash University

  1. Application of Bagging in Day-Ahead Electricity Price Forecasting and Factor Augmentation By Kadir Özen; Dilem Yıldırım
  2. Modelling uncertainty in financial tail risk: a forecasting combination and weighted quantile approach By Giuseppe Storti; Chao Wang
  3. Forecasting UK inflation bottom up By Joseph, Andreas; Kalamara, Eleni; Kapetanios, George; Potjagailo, Galina
  4. Forecasting Electricity Prices: Autoregressive Hybrid Nearest Neighbors (ARHNN) method By Weronika Nitka; Tomasz Serafin; Dimitrios Sotiros
  5. Nowcasting in a Pandemic using Non-Parametric Mixed Frequency VARs By Florian, Huber; Koop, Gary; Onorante, Luca; Pfarrhofer, Michael; Schreiner, Josef
  6. Recurrent Dictionary Learning for State-Space Models with an Application in Stock Forecasting By Shalini Sharma; Víctor Elvira; Emilie Chouzenoux; Angshul Majumdar
  7. Profitability Analysis in Stock Investment Using an LSTM-Based Deep Learning Model By Jaydip Sen; Abhishek Dutta; Sidra Mehtab
  8. Wavelet Denoised-ResNet CNN and LightGBM Method to Predict Forex Rate of Change By Yiqi Zhao; Matloob Khushi

  1. By: Kadir Özen (Barcelona Graduate School of Economics, Barcelona, Spain); Dilem Yıldırım (Department of Economics, Middle East Technical University, Ankara, Turkey)
    Abstract: The electricity price forecasting (EPF) is a challenging task not only because of the uncommon characteristics of electricity but also because of the existence of many potential predictors with changing predictive abilities over time. Particularly, how to account for all available factors and extract as much information as possible is the key to the production of accurate forecasts. To address this long-standing issue in a way that balances complexity and forecasting accuracy while facilitating the traceability of the predictor selection procedure, the method of Bootstrap Aggregation (bagging), which is a variant shrinkage estimation approach for the estimation of large scale models, is proposed in this paper. To forecast day-ahead electricity prices in a multivariate context for six major power markets we construct a large scale pure-price model (in addition to some stochastic models that are commonly applied in the literature) and apply the bagging approach in comparison with the popular Least Absolute Shrinkage and Selection Operator (LASSO) estimation method. Our forecasting study reveals that with its superior forecasting performance and its computationally simple algorithm, the bagging emerges as a strong competitor to the commonly applied LASSO approach for the short-term EPF. Further analysis for the variable selection for the bagging and LASSO approaches suggests that the differentiation in the forecast performances of two approaches might be due to, inter alia, their structural differences in the explanatory variables selection process. Moreover, to account for the intraday hourly dependencies of day-ahead electricity prices, all our models are augmented with latent factors, and a substantial improvement is observed only in the forecasts from models covering a relatively limited number of predictors, while almost no improvement is obtained in the forecasts from the large scale model estimated through LASSO and bagging techniques.
    Keywords: Bagging, Shrinkage methods, Electricity price forecasting, Multivariate modeling, Forecast encompassing, Factor models
    JEL: C22 C38 C51 C53 Q47
    Date: 2021–04
  2. By: Giuseppe Storti; Chao Wang
    Abstract: A novel forecasting combination and weighted quantile based tail risk forecasting framework is proposed, aiming to reduce the impact of modelling uncertainty in financial tail risk forecasting. The proposed approach is based on a two-step estimation procedure. The first step involves the combination of Value-at-Risk (VaR) forecasts at a grid of different quantile levels. A range of parametric and semi-parametric models is selected as the model universe which is incorporated in the forecasting combination procedure. The quantile forecasting combination weights are estimated by optimizing the quantile loss. In the second step, the Expected Shortfall (ES) is computed as a weighted average of combined quantiles. The quantiles weighting structure used to generate the ES forecast is determined by minimizing a strictly consistent joint VaR and ES loss function of the Fissler-Ziegel class. The proposed framework is applied to six stock market indices and its forecasting performance is compared to each individual model in the model universe and a simple average approach. The forecasting results based on a number of evaluations support the proposed framework.
    Date: 2021–04
  3. By: Joseph, Andreas (Bank of England); Kalamara, Eleni (King’s College London); Kapetanios, George (King’s College London); Potjagailo, Galina (Bank of England)
    Abstract: We forecast CPI inflation in the United Kingdom up to one year ahead using a large set of monthly disaggregated CPI item series combined with a wide set of forecasting tools, including dimensionality reduction techniques, shrinkage methods and non-linear machine learning models. We find that exploiting CPI item series over the period 2011–19 yields strong improvements in forecasting UK inflation against an autoregressive benchmark, above and beyond the gains from macroeconomic predictors. Ridge regression and other shrinkage methods perform best across specifications that include item-level data, yielding gains in relative forecast accuracy of up to 70% at the one-year horizon. Our results suggests that the combination of a large and relevant information set combined with efficient penalisation is key for good forecasting performance for this problem. We also provide a model-agnostic approach to address the general problem of model interpretability in high-dimensional settings based on model Shapley values, partial re-aggregation and statistical testing. This allows us to identify CPI divisions that consistently drive aggregate inflation forecasts across models and specifications, as well as to assess model differences going beyond forecast accuracy.
    Keywords: Inflation; forecasting; machine learning; state space models; CPI disaggregated data; Shapley values
    JEL: C32 C45 C53 C55 E37
    Date: 2021–03–26
  4. By: Weronika Nitka; Tomasz Serafin; Dimitrios Sotiros
    Abstract: The ongoing reshape of electricity markets has significantly stimulated electricity trading. Limitations in storing electricity as well as on-the-fly changes in demand and supply dynamics, have led price forecasts to be a fundamental aspect of traders' economic stability and growth. In this perspective, there is a broad literature that focuses on developing methods and techniques to forecast electricity prices. In this paper, we develop a new hybrid method, called ARHNN, for electricity price forecasting (EPF) in day-ahead markets. A well performing autoregressive model, with exogenous variables, is the main forecasting instrument in our method. Contrarily to the traditional statistical approaches, in which the calibration sample consists of the most recent and successive observations, we employ the k-nearest neighbors (k-NN) instance-based learning algorithm and we select the calibration sample based on a similarity (distance) measure over a subset of the autoregressive model's variables. The optimal levels of the k-NN parameter are identified during the validation period in a way that the forecasting error is minimized. We apply our method in the EPEX SPOT market in Germany. The results reveal a significant improvement in accuracy compared to commonly used approaches.
    Keywords: Electricity price forecasting; Day-ahead market; ARX; k-nearest neighbors
    JEL: C22 C32 C51 C53 Q41 Q47
    Date: 2021–04–10
  5. By: Florian, Huber (University of Salzburg); Koop, Gary (University of Strathclyde); Onorante, Luca (European Commission); Pfarrhofer, Michael (University of Salzburg); Schreiner, Josef (Oesterreichische Nationalbank)
    Abstract: This paper develops Bayesian econometric methods for posterior inference in non-parametric mixed frequency VARs using additive regression trees. We argue that regression tree models are ideally suited for macroeconomic nowcasting in the face of extreme observations, for instance those produced by the COVID-19 pandemic of 2020. This is due to their flexibility and ability to model outliers. In an application involving four major euro area countries, we find substantial improvements in nowcasting performance relative to a linear mixed frequency VAR.
    Keywords: Regression tree models, Bayesian, macroeconomic forecasting, vector autoregressions
    JEL: C11 C32 C53 E37
    Date: 2021–03
  6. By: Shalini Sharma (IIIT-Delhi - Indraprastha Institute of Information Technology [New Delhi]); Víctor Elvira (School of Mathematics - University of Edinburgh - University of Edinburgh); Emilie Chouzenoux (OPIS - OPtimisation Imagerie et Santé - CVN - Centre de vision numérique - CentraleSupélec - Université Paris-Saclay - Inria - Institut National de Recherche en Informatique et en Automatique - Inria Saclay - Ile de France - Inria - Institut National de Recherche en Informatique et en Automatique); Angshul Majumdar (IIIT-Delhi - Indraprastha Institute of Information Technology [New Delhi])
    Abstract: In this work, we introduce a new modeling and inferential tool for dynamical processing of time series. The approach is called recurrent dictionary learning (RDL). The proposed model reads as a linear Gaussian Markovian state-space model involving two linear operators, the state evolution and the observation matrices, that we assumed to be unknown. These two unknown operators (that can be seen interpreted as dictionaries) and the sequence of hidden states are jointly learnt via an expectation-maximization algorithm. The RDL model gathers several advantages, namely online processing, probabilistic inference, and a high model expressiveness which is usually typical of neural networks. RDL is particularly well suited for stock forecasting. Its performance is illustrated on two problems: next day forecasting (regression problem) and next day trading (classification problem), given past stock market observations. Experimental results show that our proposed method excels over state-of-the-art stock analysis models such as CNN-TA, MFNN, and LSTM.
    Keywords: Stock Forecasting,Recurrent dictionary learning,Kalman filter,expectation-minimization,dynamical modeling,uncertainty quantification
    Date: 2021
  7. By: Jaydip Sen; Abhishek Dutta; Sidra Mehtab
    Abstract: Designing robust systems for precise prediction of future prices of stocks has always been considered a very challenging research problem. Even more challenging is to build a system for constructing an optimum portfolio of stocks based on the forecasted future stock prices. We present a deep learning-based regression model built on a long-and-short-term memory network (LSTM) network that automatically scraps the web and extracts historical stock prices based on a stock's ticker name for a specified pair of start and end dates, and forecasts the future stock prices. We deploy the model on 75 significant stocks chosen from 15 critical sectors of the Indian stock market. For each of the stocks, the model is evaluated for its forecast accuracy. Moreover, the predicted values of the stock prices are used as the basis for investment decisions, and the returns on the investments are computed. Extensive results are presented on the performance of the model. The analysis of the results demonstrates the efficacy and effectiveness of the system and enables us to compare the profitability of the sectors from the point of view of the investors in the stock market.
    Date: 2021–04
  8. By: Yiqi Zhao; Matloob Khushi
    Abstract: Foreign Exchange (Forex) is the largest financial market in the world. The daily trading volume of the Forex market is much higher than that of stock and futures markets. Therefore, it is of great significance for investors to establish a foreign exchange forecast model. In this paper, we propose a Wavelet Denoised-ResNet with LightGBM model to predict the rate of change of Forex price after five time intervals to allow enough time to execute trades. All the prices are denoised by wavelet transform, and a matrix of 30 time intervals is formed by calculating technical indicators. Image features are obtained by feeding the maxtrix into a ResNet. Finally, the technical indicators and image features are fed to LightGBM. Our experiments on 5-minutes USDJPY demonstrate that the model outperforms baseline modles with MAE: 0.240977x10EXP-3 MSE: 0.156x10EXP-6 and RMSE: 0.395185x10EXP-3. An accurate price prediction after 25 minutes in future provides a window of opportunity for hedge funds algorithm trading. The code is available from
    Date: 2021–01

This nep-for issue is ©2021 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.