nep-for New Economics Papers
on Forecasting
Issue of 2020‒04‒20
seven papers chosen by
Rob J Hyndman
Monash University

  1. Forecasts with Bayesian vector autoregressions under real time conditions By Michael Pfarrhofer
  2. Financial Market Trend Forecasting and Performance Analysis Using LSTM By Jonghyeon Min
  3. Forecasting exchange rates of major currencies with long maturity forward rates By Zsolt Darvas; Zoltán Schepp
  4. Machine Learning Algorithms for Financial Asset Price Forecasting By Philip Ndikum
  5. The effects of targeting predictors in a random forest regression model By Daniel Borup; Bent Jesper Christensen; Nicolaj N{\o}rgaard M\"uhlbach; Mikkel Slot Nielsen
  6. A Deep Reinforcement Learning Framework for Continuous Intraday Market Bidding By Ioannis Boukas; Damien Ernst; Thibaut Th\'eate; Adrien Bolland; Alexandre Huynen; Martin Buchwald; Christelle Wynants; Bertrand Corn\'elusse
  7. First in, First out: Econometric Modelling of UK Annual CO_2 Emissions, 1860–2017 By David F. Hendry

  1. By: Michael Pfarrhofer
    Abstract: This paper investigates the sensitivity of forecast performance measures to taking a real time versus pseudo out-of-sample perspective. We use monthly vintages for the United States (US) and the Euro Area (EA) and estimate a set of vector autoregressive (VAR) models of different sizes with constant and time-varying parameters (TVPs) and stochastic volatility (SV). Our results suggest differences in the relative ordering of model performance for point and density forecasts depending on whether real time data or truncated final vintages in pseudo out-of-sample simulations are used for evaluating forecasts. No clearly superior specification for the US or the EA across variable types and forecast horizons can be identified, although larger models featuring TVPs appear to be affected the least by missing values and data revisions. We identify substantial differences in performance metrics with respect to whether forecasts are produced for the US or the EA.
    Date: 2020–04
  2. By: Jonghyeon Min
    Abstract: The financial market trend forecasting method is emerging as a hot topic in financial markets today. Many challenges still currently remain, and various researches related thereto have been actively conducted. Especially, recent research of neural network-based financial market trend prediction has attracted much attention. However, previous researches do not deal with the financial market forecasting method based on LSTM which has good performance in time series data. There is also a lack of comparative analysis in the performance of neural network-based prediction techniques and traditional prediction techniques. In this paper, we propose a financial market trend forecasting method using LSTM and analyze the performance with existing financial market trend forecasting methods through experiments. This method prepares the input data set through the data preprocessing process so as to reflect all the fundamental data, technical data and qualitative data used in the financial data analysis, and makes comprehensive financial market analysis through LSTM. In this paper, we experiment and compare performances of existing financial market trend forecasting models, and performance according to the financial market environment. In addition, we implement the proposed method using open sources and platform and forecast financial market trends using various financial data indicators.
    Date: 2020–03
  3. By: Zsolt Darvas; Zoltán Schepp
    Abstract: The theoretical derivation of our forecasting equation is consistent with the monetary model of exchange rates. Our model outperforms the random walk in out-of-sample forecasting of twelve major currency pairs in both short and long horizons forecasts for the 1990-2020 period. The results are robust for all sub-periods with the exception of years around the collapse of Lehman Brothers in September 2008. Our results are robust to alternative model specifications,...
    Date: 2020–04
  4. By: Philip Ndikum
    Abstract: This research paper explores the performance of Machine Learning (ML) algorithms and techniques that can be used for financial asset price forecasting. The prediction and forecasting of asset prices and returns remains one of the most challenging and exciting problems for quantitative finance and practitioners alike. The massive increase in data generated and captured in recent years presents an opportunity to leverage Machine Learning algorithms. This study directly compares and contrasts state-of-the-art implementations of modern Machine Learning algorithms on high performance computing (HPC) infrastructures versus the traditional and highly popular Capital Asset Pricing Model (CAPM) on U.S equities data. The implemented Machine Learning models - trained on time series data for an entire stock universe (in addition to exogenous macroeconomic variables) significantly outperform the CAPM on out-of-sample (OOS) test data.
    Date: 2020–03
  5. By: Daniel Borup; Bent Jesper Christensen; Nicolaj N{\o}rgaard M\"uhlbach; Mikkel Slot Nielsen
    Abstract: The random forest regression (RF) has become an extremely popular tool to analyze high-dimensional data. Nonetheless, it has been argued that its benefits are lessened in sparse high-dimensional settings due to the presence of weak predictors and an initial dimension reduction (targeting) step prior to estimation is required. We show theoretically that, in high-dimensional settings with limited signal, proper targeting is an important complement to RF's feature sampling by controlling the probability of placing splits along strong predictors. This is supported by simulations with representable finite samples. Moreover, we quantify the immediate gain from targeting in terms of increased strength of individual trees. Our conclusions are elaborated by a broad set of applications within macroeconomics and finance. These show that the inherent bias-variance trade-off implied by targeting, due to increased tree correlation, is balanced at a medium level, selecting the best 10-30\% of commonly applied predictors. The applications consolidate that improvements from the targeted RF over the ordinary RF can be significant, particularly in long-horizon forecasting, and both in expansions and recessions.
    Date: 2020–04
  6. By: Ioannis Boukas; Damien Ernst; Thibaut Th\'eate; Adrien Bolland; Alexandre Huynen; Martin Buchwald; Christelle Wynants; Bertrand Corn\'elusse
    Abstract: The large integration of variable energy resources is expected to shift a large part of the energy exchanges closer to real-time, where more accurate forecasts are available. In this context, the short-term electricity markets and in particular the intraday market are considered a suitable trading floor for these exchanges to occur. A key component for the successful renewable energy sources integration is the usage of energy storage. In this paper, we propose a novel modelling framework for the strategic participation of energy storage in the European continuous intraday market where exchanges occur through a centralized order book. The goal of the storage device operator is the maximization of the profits received over the entire trading horizon, while taking into account the operational constraints of the unit. The sequential decision-making problem of trading in the intraday market is modelled as a Markov Decision Process. An asynchronous distributed version of the fitted Q iteration algorithm is chosen for solving this problem due to its sample efficiency. The large and variable number of the existing orders in the order book motivates the use of high-level actions and an alternative state representation. Historical data are used for the generation of a large number of artificial trajectories in order to address exploration issues during the learning process. The resulting policy is back-tested and compared against a benchmark strategy that is the current industrial standard. Results indicate that the agent converges to a policy that achieves in average higher total revenues than the benchmark strategy.
    Date: 2020–04
  7. By: David F. Hendry (Institute for New Economic Thinking at the Oxford Martin School and Climate Econometrics, Nuffield College, University of Oxford)
    Abstract: The United Kingdom was the first country into the Industrial Revolution in the mid-18th Century. 250 years later, real income levels in the UK are about 7-10 fold higher per capita, even greater elsewhere, many killer diseases have been tamed, and longevity has approximately doubled. However, such beneficial developments have led to a global explosion in anthropogenic emissions of greenhouse gases. Following the Climate Change Act of 2008, the UK is now one of the first countries out, with annual CO_2 emissions per capita below 1860’s levels. We develop an econometric model of its highly non-stationary emissions process over the last 150 years, confirming the key roles of reduced coal use and of the capital stock, which embodies the vintage of technology at its construction. Major shifts and outliers must be handled to develop a viable model, and the advantages of doing so are detecting the impacts of important policies and improved forecasts. Large reductions in all CO_2 sources will be required to meet the 2050 target of an 80% reduction from 1970 levels, and their near elimination for a net-zero level.
    Keywords: UK CO2 Emissions; Model Selection; Saturation Estimation; Autometrics; Climate Change Act; Climate Policy Implications.
    JEL: C51 Q54
    Date: 2020–02–07

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.