nep-for New Economics Papers
on Forecasting
Issue of 2016‒05‒14
thirteen papers chosen by
Rob J Hyndman
Monash University

  1. Forecasting the USD/CNY Exchange Rate under Different Policy Regimes By Yuxuan Huang
  2. Joint Prediction Bands for Macroeconomic Risk Management By Farooq Akram; Andrew Binning; Junior Maih
  3. Forecasting GDP with global components. This time is different By Hilde C. Bjornland; Francesco Ravazzolo; Leif Anders Thorsrud
  4. Updated Reference Forecasts for Global CO2 Emissions from Fossil-Fuel Consumption By José M. Belbute; Alfredo Marvão Pereira
  5. Forecasting Inflation using Functional Time Series Analysis By Zafar, Raja Fawad; Qayyum, Abdul; Ghouri, Saghir Pervaiz
  6. A Practical Note on Predictive Analytics Usage in Marketing Applications By Banerjee, Arindam; Banerjee, Tanushri
  7. Improving Markov switching models using realized variance By Liu, Jia; Maheu, John M
  8. CBO’s Revenue Forecasting Record By Congressional Budget Office
  9. Are experts’ probabilistic forecasts similar to the NBP projections? By Halina Kowalczyk; Ewa Stanisławska
  10. How did farmers act? An ex-post validation of normative and positive mathematical programming for an agent-based sector model By Mack, Gabriele; Ferjani, Ali; Mohring, Anke; Zimmerman, Albert; Mann, Stefan
  11. Modelling and Forecasting Mortgage Delinquency and Foreclosure in the UK By Aron, Janine; Muellbauer, John
  13. Do Terror Attacks Predict Gold Returns? Evidence from a Quantile-Predictive-Regression Approach By Rangan Gupta; Anandamayee Majumdar; Christian Pierdzioch; Mark Wohar

  1. By: Yuxuan Huang (The George Washington University)
    Abstract: The USD/CNY exchange rate exhibits very different pattern in different periods as it changes wildly from one period to another according to the economic reforms and policies. This paper compares the performance of six different forecasting models of USD/CNY exchange rate under three different forecast scenarios from 2005-2015. In particular, the paper focuses in answering the following questions. (i) Do models' forecast performance change when the marketization level changes? (ii) Which model has the best forecast when the regimes change? (iii) Can forecasting robustifications help? The forecast results show that models incorporates economic fundamentals perform better in less regulated periods when the exchange rate can oat more freely. For the forecast experiments with breaks in the forecast origin, the exchange rate CVAR model perform the best before robustifications. In most cases, the intercept-correction and doubledifference device improve the forecast performance in both dynamic forecast and one-step forecast. Different models seem to do well under different forecast scenario after applying the robust devices.
    Keywords: Exchange rates; Forecasting; US Dollar; Chinese Yuan
    JEL: F31 F37 C53
    Date: 2016–01
  2. By: Farooq Akram; Andrew Binning; Junior Maih
    Abstract: In this paper we address the issue of assessing and communicating the joint probabilities implied by density forecasts from multivariate time series models. We focus our attention in three areas. First, we investigate a new method of producing fan charts that better communicates the uncertainty present in forecasts from multivariate time series models. Second, we suggest a new measure for assessing the plausibility of non-central point forecasts. And third, we describe how to use the density forecasts from a multivariate time series model to assess the probability of a set of future events occurring. An additional novelty of this paper is our use of a regime-switching DSGE model with an occasionally binding zero lower bound constraint, estimated on US data, to produce the density forecasts. The tools we offer will allow practitioners to better assess and communicate joint forecast probabilities, a criticism that has been leveled at central bank communications.
    Keywords: Monetary Policy, Fan charts, DSGE, Zero Lower Bound, Regime-switching, Bayesian Estimation
    Date: 2016–05
  3. By: Hilde C. Bjornland; Francesco Ravazzolo; Leif Anders Thorsrud
    Abstract: We examine whether knowledge of in-sample co-movement across countries can be used in a more systematic way to improve forecast accuracy at the national level. In particular, we ask if a model with common international business cycle factors adds marginal predictive power compared to a domestic alternative? To answer this question we use a Dynamic Factor Model (DFM) and run an out-of-sample forecasting experiment. Our results show that exploiting the informational content in a common global business cycle factor improves forecast accuracy in terms of both point and density forecast evaluation across a large panel of countries. We also document that the Great Recession has a huge impact on this result, causing a clear preference shift towards the model including a common global factor. However, this time is different also in other respects. On longer forecasting horizons the performance of the DFM deteriorates substantially in the aftermath of the Great Recession.
    Keywords: Bayesian Dynamic Factor Model (BDFM), forecasting, model uncertainty and global factors
    JEL: C11 C53 C55 F17
    Date: 2016–05
  4. By: José M. Belbute (Department of Economics, University of Évora, Portugal); Alfredo Marvão Pereira (Department of Economics, The College of William and Mary)
    Abstract: We provide alternative reference forecasts for global CO2 emissionsbased on an ARFIMA model estimated with annual data from 1750 to 2014. These forecasts are free from additional assumptions on demographic and economic variables that are commonly used in reference forecasts, as they only rely on the properties of the underlying stochastic process for CO2emissions, as well ason all the observed information it incorporates. In this sense, these forecasts are morebased on fundamentals. Our reference forecast suggests that in 2030, 2040 and 2050, in the absence of any structural changes of any type, CO2would likely be at about 23.1%, 29.1% and 33.7% above 2010 emission levels, respectively. These values are clearly below the levels proposed by other reference scenarios available in the literature.This is important, as it suggests that the ongoing policy goals are actually within much closer reach than what is implied by the standard CO2reference emission scenarios. Having lower and more realistic reference emissions projections not only gives a truer assessment of the policy efforts that are needed,but also highlights the lower costs involved in mitigation efforts, thereby maximizing the likelihood of more widespread energy and environmental policy efforts.
    Keywords: Forecasting, reference scenario, CO2 emissions, long memory, ARFIMA
    JEL: C22 C53 O13 Q47 Q54
    Date: 2016–05–06
  5. By: Zafar, Raja Fawad; Qayyum, Abdul; Ghouri, Saghir Pervaiz
    Abstract: In present study we model the data using Functional Time series Analysis (FTSA). The method is basically univariate, so to check its efficiency we compared it with seasonal ARIMA models. We have used data sets of monthly frequency from 2002-2011 to forecast Consumer Price Index (CPI) of Pakistan. We withhold some data of last year (i.e. of 2011) and based on remaining year (2002-2010) we fitted model and forecasted the values of monthly CPI. Our study compares the performance of FTSA model and ARIMA model using the test data of 2011. Comparison based on forecast evaluation criteria’s and forecasted value of 2011, indicates that FTSA model using CPI general data outperforms SARIMA models
    Keywords: Forecasting, Inflation, SARIMA, FTSA
    JEL: C22 C53
    Date: 2015–03–20
  6. By: Banerjee, Arindam; Banerjee, Tanushri
    Abstract: Most Predictive Analytics discussions focus on methods that can be used for better quality prediction in a particular context. Realizing that the possibility of perfect prediction is a near impossibility, practitioners looking to support their futuristic initiatives wonder, what is a suitable model for their use. In other words, if all prediction models are imperfect (have leakage) how much of this imperfection can be tolerated and yet better decisions can be taken with model output. This paper is an attempt to provide a simplified approach to this practical problem of evaluating model performance taking account of the decision context. Two scenarios are discussed; a) a classification problem often used for profiling customers into segments and, b) a volume forecasting problem. In both cases, the leakage is defined (misclassification or uncertainty band) and their impact (adverse) on the subsequent decision is identified. Contextual dimensions that have an impact on the quality of the decision and the scope to alleviate the problem are also discussed.
  7. By: Liu, Jia; Maheu, John M
    Abstract: This paper proposes a class of models that jointly model returns and ex-post variance measures under a Markov switching framework. Both univariate and multivariate return versions of the model are introduced. Bayesian estimation can be conducted under a fixed dimension state space or an infinite one. The proposed models can be seen as nonlinear common factor models subject to Markov switching and are able to exploit the information content in both returns and ex-post volatility measures. Applications to U.S. equity returns and foreign exchange rates compare the proposed models to existing alternatives. The empirical results show that the joint models improve density forecasts for returns and point predictions of return variance. The joint Markov switching models can increase the precision of parameter estimates and sharpen the inference of the latent state variable.
    Keywords: infinite hidden Markov model, realized covariance, density forecast, MCMC
    JEL: C11 C32 C51 C58 G1
    Date: 2015–09–01
  8. By: Congressional Budget Office
    Abstract: This report examines the accuracy of CBO’s revenue projections since 1982, the earliest year for which the information necessary to assess the forecasts is available. On average, the agency’s projections have been a bit too high—more so for projections spanning six years than for those spanning two—owing mostly to the difficulty of predicting when economic downturns will occur. The overall accuracy of CBO’s revenue projections has been similar to that of the projections of other government agencies.
    JEL: H20 C53
    Date: 2015–11–10
  9. By: Halina Kowalczyk; Ewa Stanisławska
    Abstract: We assess similarity of the Polish central bank’s forecasts published in Inflation Reports and economic experts’ forecasts (from NBP Survey of Professional Forecasters), an important issue in monetary policy. Contrary to other studies which use point forecasts, we are interested is comparing whole forecasts’ distributions. We are especially interested whether the SPF experts mirror the NBP projections. For this purpose, we propose employing methods based on distance between distributions. Unfortunately, substantial heterogeneity of forecasts, as well as short and atypical period analyzed, limit drawing firm conclusions with this respect.
    Keywords: survey data, fan charts, probabilistic forecasts, inflation forecasts, GDP growth forecasts, distribution similarity
    JEL: D83 D84 E37
    Date: 2016
  10. By: Mack, Gabriele; Ferjani, Ali; Mohring, Anke; Zimmerman, Albert; Mann, Stefan
    Abstract: This study evaluates normative (NMP) and positive (PMP) mathematical programming methods for the recursive dynamic agent-based sector model SWISSland, which determines production decisions for 3400 farm-level models for the ex-post period 2005 to 2012. This study clearly shows that PMP for crop production activities improves the forecasting performance of farm based agent-based models compared to NMP. It also shows that combining PMP and NMP could be a suitable approach for agent-based sector models. For short-term forecast PMP for all production activities and PMP combined with NMP lead to similar results. The results either show that PMP calibration based on revenues and PMP calibration based on the entropy approach lead to similar results. By combining PMP with NMP some limitations of PMP could be reduced. In branches where the adoption of new production activities is expected due to market, the NMP approach could be an appropriate solution.
    Keywords: agent-based sector model, positive mathematical programming, ex-post validation, Agricultural and Food Policy, Agricultural Finance,
    Date: 2015
  11. By: Aron, Janine; Muellbauer, John
    Abstract: In the absence of micro-data in the public domain, new aggregate models for the UK's mortgage repossessions and arrears are estimated using quarterly data over 1983-2014, motivated by a conceptual double trigger frame framework for foreclosures and payment delinquencies. An innovation to improve on the flawed but widespread use of loan-to-value measures, is to estimate difficult-to-observe variations in loan quality and access to refinancing, and shifts in lenders' forbearance policy, by common latent variables in a system of equations for arrears and repossessions. We introduce, for the first time in the literature, a theory-justified estimate of the proportion of mortgages in negative equity as a key driver of aggregate repossessions and arrears. This is based on an average debt-equity ratio, corrected for regional deviations, and uses a functional form for the distribution of the debt-equity ratio checked on Irish micro-data from the Bank of Ireland, and Bank of England snapshots of negative equity. We systematically address serious measurement bias in the 'months-in-arrears' measures, neglected in previous UK studies. Highly significant effects on aggregate rates of repossessions and arrears are found for the aggregate debt-service ratio, the proportion of mortgages in negative equity and the unemployment rate. Economic forecast scenarios to 2020 highlight risks faced by the UK and its mortgage lenders, illustrating the usefulness of the approach for bank stress-testing. For macroeconomics, our model traces an important part of the financial accelerator: the feedback from the housing market to bad loans and hence banks' ability to extend credit.
    Keywords: credit risk stress testing; foreclosures; latent variables model; mortgage arrears; mortgage payment delinquencies; mortgage repossessions
    JEL: C51 C53 E27 G17 G21 G28 R21 R28
    Date: 2016–04
  12. By: Ulrich Heilemann (Universität Leipzig); Susanne Schnorr-Bäcker (Statistisches Bundesamt)
    Abstract: Given that the Great Recession in Germany was neither predicted nor identified at the time, this paper examines whether data available could have helped to predict or identify the crisis in real time. We inspect forecasts published during April–December 2008 by 12 major institutions, for available data: real-time data from official statistics for Germany and the European Union, major surveys, and indicators. Although annual real GDP forecasts for 2008 were rather accurate, forecasters failed to observe the onset of the recession in Q2 2008, though from May onward, an increasing amount of data—neither ambiguous nor misleading—indicated that the economy was in recession or would likely enter one soon. Nevertheless, forecasters recognised the recession only in mid-November, when the country was already seven months into the recession, thereby confirming forecasters’ ‘low priors about the likelihood of a recession’.
    Keywords: Forecast accuracy; Great Recession; real-time analysis; data processing
    JEL: C53 E32 E37
    Date: 2016–02
  13. By: Rangan Gupta (Department of Economics, University of Pretoria); Anandamayee Majumdar (Center for Advanced Statistics and Econometrics, Soochow University, China); Christian Pierdzioch (Department of Economics, Helmut Schmidt University, Germany); Mark Wohar (Department of Economics, University of Nebraska-Omaha, USA and Loughborough University, UK)
    Abstract: Much significant research has been done to study how terror attacks affect financial markets. We contribute to this research by studying whether terror attacks, in addition to standard predictors considered in earlier research, help to predict gold returns. To this end, we use a Quantile-Predictive-Regression (QPR) approach that accounts for model uncertainty and model instability. We find that terror attacks have predictive value for the lower and especially for the upper quantiles of the conditional distribution of gold returns.
    Keywords: Gold returns, Terror attacks, Forecasting model, Quantile regression
    JEL: C22 C53 Q02
    Date: 2016–03

This nep-for issue is ©2016 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.