nep-for New Economics Papers
on Forecasting
Issue of 2018‒08‒13
twelve papers chosen by
Rob J Hyndman
Monash University

  1. NKPC-Based Inflation Forecasts with a Time-Varying Trend By Stephen McKnight; Alexander Mihailov; Fabio Rumler
  2. The Forcasting Performance of Dynamic Factor Models with Vintage Data By Di Bonaventura, Luca; Forni, Mario; Pattarin, Francesco
  3. A Machine Learning Approach to the Forecast Combination Puzzle By Antoine Mandel; Amir Sani
  4. Testing the predictability of commodity prices in stock returns: A new perspective By Afees A. Salisu; Kazeem Isah; Ibrahim D. Raheem
  5. The Role of Housing Sentiment in Forecasting US Home Sales Growth: Evidence from a Bayesian Compressed Vector Autoregressive Model By Rangan Gupta; Chi Keung Marco Lau; Vasilios Plakandaras; Wing-Keung Wong
  6. Forecasting market states By Pier Francesco Procacci; Tomaso Aste
  7. Real-time forecast combinations for the oil price By Anthony Garratt; Shaun P. Vahey; Yunyi Zhang
  8. IW Financial Expert Survey: Third Quarter 2018 By Demary, Markus
  9. Does time-variation matter in the stochastic volatility components for G7 stock returns By Afees A. Salisu; Ahamuefula Ephraim Ogbonna
  10. Periodic or Generational Actuarial Tables: Which One to Choose? By Severine Arnold (-Gaille); Anca Jijiie; Eric Jondeau; Michael Rockinger
  11. Monetary Policy Lessons from the Greenbook By Michael T. Belongia; Peter N. Ireland
  12. Analysis of shock transmissions to a small open emerging economy using a SVARMA model By Raghavan, Mala; Athanasopoulos, George

  1. By: Stephen McKnight (El Colegio de México); Alexander Mihailov (University of Reading); Fabio Rumler (Oesterreichische Nationalbank)
    Abstract: Does theory aid inflation forecasting? This paper develops a forecasting procedure based upon a generalized New Keynesian Phillips Curve that in- corporates time-varying trend inflation. Using quarterly data for the Euro Area and the United States over the period 1970-2015, we decompose infla- tion into trend and cyclical components and generate theory-implied predic- tions for both, which are recombined to obtain an overall inflation forecast. We find that our forecasting procedure outperforms in predictive accuracy the conventional random walk benchmark at all horizons considered (up to 20 quarters). Moreover, it also outperforms quantitatively the agnos- tic Atkeson-Ohanian (2001) benchmark that previous studies have found dificult to beat.
    Keywords: time-varying trend, generalized New Keynesian Phillips Curve, inflation dynamics, inflation forecasts, predictive accuracy
    JEL: C53 D43 E31 E37 F41 F47
    Date: 2018–07
  2. By: Di Bonaventura, Luca; Forni, Mario; Pattarin, Francesco
    Abstract: We present a comparative analysis of the forecasting performance of two dynamic factor models, the Stock and Watson (2002a, b) model and the Forni, Hallin, Lippi and Reichlin (2005) model, based on vintage data. Our dataset contains 107 monthly US "first release" macroeconomic and financial vintage time series, spanning the 1996:12 to 2017:6 period with monthly periodicity, extracted from the Bloomberg database†. We compute real-time one-month-ahead forecasts with both models for four key macroeconomic variables: the month-on-month change in industrial production, the unemployment rate, the core consumer price index and the ISM Purchasing Managers' Index. First, we find that both the Stock and Watson and the Forni, Hallin, Lippi and Reichlin models outperform simple autoregressions for industrial production, unemployment rate and consumer prices, but that only the first model does so for the PMI. Second, we find that neither models always outperform the other. While Forni, Hallin, Lippi and Reichlin's beats Stock and Watson's in forecasting industrial production and consumer prices, the opposite happens for the unemployment rate and the PMI.
    Keywords: Dynamic factor models; First release data; Forecasting; Forecasting Performance; Vintage data
    JEL: C01 C32 C52 C53 E27 E37
    Date: 2018–07
  3. By: Antoine Mandel (CES - Centre d'économie de la Sorbonne - UP1 - Université Panthéon-Sorbonne - CNRS - Centre National de la Recherche Scientifique); Amir Sani (CFM-Imperial Institute of Quantitative Finance - Imperial College London, CES - Centre d'économie de la Sorbonne - UP1 - Université Panthéon-Sorbonne - CNRS - Centre National de la Recherche Scientifique)
    Abstract: Forecast combination algorithms provide a robust solution to noisy data and shifting process dynamics. However in practice, sophisticated combination methods often fail to consistently outperform the simple mean combination. This "forecast combination puzzle" limits the adoption of alternative com- bination approaches and forecasting algorithms by policy-makers. Through an adaptive machine learning algorithm designed for streaming data, this pa- per proposes a novel time-varying forecast combination approach that retains distribution-free guarantees in performance while automatically adapting com- binations according to the performance of any selected combination approach or forecaster. In particular, the proposed algorithm offers policy-makers the ability to compute the worst-case loss with respect to the mean combination ex-ante, while also guaranteeing that the combination performance is never worse than this explicit guarantee. Theoretical bounds are reported with re- spect to the relative mean squared forecast error. Out-of-sample empirical performance is evaluated on the Stock and Watson seven-country dataset and the ECB Sur- vey of Professional Forecasters.
    Keywords: Forecasting,Forecast Combination Puzzle,Forecast combinations,Machine Learning,Econometrics,Apprentissage statistique,Combinaison de prédicteurs,Econométrie
    Date: 2017–04–19
  4. By: Afees A. Salisu (Department for Management of Science and Technology Development, Ton Duc Thang University, Ho Chi Minh City, Vietnam Faculty of Business Administration, Ton Duc Thang University, Ho Chi Minh City, Vietnam Centre for Econometric and Allied Research, University of Ibadan); Kazeem Isah (Centre for Econometric and Allied Research, University of Ibadan); Ibrahim D. Raheem (School of Economics, University of Kent, Canterbury, UK)
    Abstract: In this paper, we offer an alternative approach for testing the predictive power of commodity prices in stock returns using monthly data of about six decades. In the process, we account for prominent features of predictive models such as asymmetry, conditional heteroscedasticity, endogeneity, persistence, and structural breaks that may bias the forecast outcomes. Using the G7 stock exchanges, three findings are discernible from the various analyses. First, commodity prices are good predictors of stock returns both for in-sample and out-of-sample forecasts. Second, the proposed commodity-based model for stock returns that accounts for the highlighted features outperforms both the traditional predictive model as well historical average models that ignore same. Thirdly, these conclusions are robust to different components of commodity prices, multiple data samples and alternative forecast horizons.
    Keywords: Stock prices, Commodity prices, G7 countries, Asymmetry, Persistence, Endogeneity; Conditional heteroscedasticity; Structural breaks
    Date: 2018–07
  5. By: Rangan Gupta (Department of Economics, University of Pretoria, Pretoria, South Africa); Chi Keung Marco Lau (Huddersfield Business School, University of Huddersfield, Huddersfield, United Kingdom); Vasilios Plakandaras (Department of Economics, Democritus University of Thrace, Komotini, Greece); Wing-Keung Wong (Department of Finance, Fintech Center, and Big Data Research Center, Asia University; Department of Medical Research, China Medical University Hospital, Taiwan; Department of Economics and Finance, Hang Seng Management College, Hong Kong, China; Department of Economics, Lingnan University, Hong Kong, China)
    Abstract: Accurate forecasts of home sales can provide valuable information for not only policymakers, but also financial institutions and real estate professionals. Against this backdrop, the objective of our paper is to analyze the role of consumers’ home buying attitudes in forecasting quarterly US home sales growth. Our results show that the home sentiment index in standard classical and Minnesota prior-based Bayesian VARs fail to add to the forecasting accuracy of the growth of home sales derived from standard economic variables already included in the models. However, when shrinkage is achieved by compressing the data using a Bayesian compressed VAR (instead of the parameters as in the BVAR), growth of US home sales can be forecasted more accurately, with the housing market sentiment improving the accuracy of the forecasts relative to the information contained in economic variables only.
    Keywords: Home Sales, Housing Sentiment, Classical and Bayesian Vector Autoregressive Models
    JEL: C32 R31
    Date: 2018–07
  6. By: Pier Francesco Procacci; Tomaso Aste
    Abstract: We propose a novel methodology to define, analyse and forecast market states. In our approach market states are identified by a reference sparse precision matrix and a vector of expectation values. In our procedure each multivariate observation is associated to a given market state accordingly to a penalized likelihood maximization. The procedure is made computationally very efficient and can be used with a large number of assets. We demonstrate that this procedure successfully classifies different states of the markets in an unsupervised manner. In particular, we describe an experiment with one hundred log-returns and two states in which the methodology automatically associates one state to periods with average positive returns and the other state to periods with average negative returns, therefore discovering spontaneously the common classification of `bull' and `bear' markets. In another experiment, with again one hundred log-returns and two states, we demonstrate that this procedure can be efficiently used to forecast off-sample future market states with significant prediction accuracy. This methodology opens the way to a range of applications in risk management and trading strategies where the correlation structure plays a central role.
    Date: 2018–07
  7. By: Anthony Garratt; Shaun P. Vahey; Yunyi Zhang
    Abstract: Baumeister and Kilian (2015) combine forecasts from six empirical models to predict real oil prices. In this paper, we broadly reproduce their main economic findings, employing their preferred measures of the real oil price and similar real-time variables. Mindful of the importance of Brent crude oil as a global price benchmark, we extend consideration to the North Sea based measure and update the evaluation sample to 2017:12. We model the oil price futures curve using a factor-based Nelson-Siegel specification to fill in missing values of oil price futures in the source data. We find that the combined forecasts for Brent are as effective as for other oil price measures. The extended sample using the oil price measures adopted by Baumeister and Kilian (2015) yields similar results to those reported in their paper. And the futures-based model improves forecast accuracy at longer horizon forecasts. The real-time data set is available for download from
    Keywords: Real oil price forecasting, Brent crude oil, Forecast combination
    Date: 2018–08
  8. By: Demary, Markus
    Abstract: The experts of the IW Financial Expert Survey predict a steeper yield curve with a larger increase in long-term than in short-term interest rates by the end of 2018. Moreover, the average forecasts indicate higher stock market indices, a mild depreciation of the Euro vis-à-vis the US-Dollar, but a larger drop in oil prices by the end of the fourth quarter of 2018. However, despite the expectation of higher interest rates, the short-term interest rate is predicted to remain in negative territory. The 3-month Euribor is, on average, expected to reach -0.31 percent at the end of the fourth quarter of 2018, while the yield on German government bonds with 10-year maturity is expected to reach 0.75 percent by then. Stock markets are, on average, expected to increase by 6.8 percent (Stoxx 50) and 9.4 percent (DAX 30) until the end of the year 2018. During that same period, the experts predict a mild depreciation of the Euro by 0.89 percent vis-à-vis the US Dollar, while oil prices are expected to drop by 9.0 percent. The expectation of an increase in the long rate and a slight increase in the short rate, together with the expected delayed monetary tightening of the ECB, hint at a financial market outlook characterised by a cautious approach to monetary normalisation. In this cautious approach, the ECB lets the market determine the first increases in long-term interest rates before it stops intervening at the long end of the yield curve, while keeping the short end of the yield curve lower. This cautious approach to monetary policy normalization is reflected in the projection of the yield curve. Moreover, the experts expect that the development of the Euro and the development of oil prices as well as the development of the stock market will support the ECB's cautious approach to monetary normalization instead of forcing a faster exit from low interest rates. Each quarter we ask the participants alternating questions on current topics. At this time, we were interested in their opinion on the proposed Euro area reforms. The majority of the surveyed experts is of the opinion that a European monetary fund is needed in the Euro area. Moreover, a majority of the same size thinks that macroprudential policies are worthwhile for stabilizing the Euro area. In addition to that, the same majority sees the need for insolvency proce-dures for states. A small number of experts only sees a European finance minister and the introduction of sovereign-bond-backed securities as worthwhile. These proposals have the highest number of experts who think of them as counter-productive. The evaluation of the forecasting performance of the latest forecasts yielded the result that Deutsche Bank performed best in predicting trends in the long-term ranking, which covers all forecasts from March 2015 to March 2018. The experts of Deutsche Bank also performed best in the short-term ranking, which covers the surveys from December 2017 and March 2018 for the 3-months ahead prediction and the survey for December 2017 for the 6-month forecasts. When it comes to point prediction, in the long-term evaluation of the period running from March 2015 to March 2018, the experts of National-Bank performed best in predicting all indicators, while the Commerzbank experts produced the most precise point forecasts for all indicators for the short-term evaluation period.
    JEL: G12 G17
    Date: 2018
  9. By: Afees A. Salisu (Department for Management of Science and Technology Development, Ton Duc Thang University, Ho Chi Minh City, Vietnam Faculty of Business Administration, Ton Duc Thang University, Ho Chi Minh City, Vietnam Centre for Econometric and Allied Research, University of Ibadan); Ahamuefula Ephraim Ogbonna (Centre for Econometric and Allied Research, University of Ibadan Department of Statistics, University of Ibadan, Ibadan, Nigeria)
    Abstract: This study empirically tests for time variation in the stochastic volatility (SV) components for the G7 stock returns. The time variation in both trend and transitory components of the SV is tested separately and jointly using the unobserved component model and following the approach developed by Chan (2018). Consequently, the computed Bayes factor obtained from the SavageDickey density ratio, which circumvents the computation of marginal likelihood, is used to adjudge the performance of each restricted time varying stochastic volatility model without the trend and transitory components against the unrestricted model that allows for same. The empirical evidence supports time variation in the transitory component of SV while the trend component is found to be relatively constant over time. These empirical estimates are not sensitive to data frequency.
    Keywords: Bayesian; Bayes factor; Transitory component; Trend component; Unobserved Component Model
    JEL: C11 C32 C53 E37 G17
    Date: 2018–07
  10. By: Severine Arnold (-Gaille) (University of Lausanne); Anca Jijiie (Faculty of Business and Economics); Eric Jondeau (University of Lausanne and Swiss Finance Institute); Michael Rockinger (University of Lausanne, Centre for Economic Policy Research (CEPR), and Swiss Finance Institute)
    Abstract: The increase in life expectancy over the past several decades has been impressive and represents a key challenge for institutions that provide life insurance products. Indeed, when a new actuarial table is released with updated survival and death rates, such institutions need to update the amount of mathematical reserve that they need to set aside to guarantee the future payments of their annuities. As mortality forecasting techniques are currently well developed, it is relatively easy to forecast mortality over several decades and to directly use these forecast rates in the determination of the mathematical reserve needed to guarantee annuity payments. Future mortality evolution is then directly incorporated into the liabilities valuation of an institution, and it is thus commonly believed that such liabilities should not require much updating when a new actuarial table is released. In this paper, we demonstrate that contrary to this common belief, institutions that use generational tables (namely, tables including future mortality evolution) will most likely need to make more important adjustments (positive or negative) to their liabilities than will institutions using periodic (static) tables whenever a new table is released. By using three very different models to project mortality, we demonstrate that our findings are inherent in the required long horizons of the forecasts needed in the generational approach, with the uncertainty surrounding the forecast values increasing with the horizon. Therefore, generational tables may introduce more instability in a pension institution’s accounts than periodic tables.
    Keywords: Mortality rates, Periodic actuarial tables, Generational actuarial tables, Life expectancy, Mathematical reserve, Mortality forecasts
    Date: 2018–01
  11. By: Michael T. Belongia (University of Mississippi); Peter N. Ireland (Boston College)
    Abstract: From 1987 through 2012, the Federal Open Market Committee appears to have set its federal funds rate target with reference to Greenbook forecasts of the output gap and inflation and to have made further adjustments to the funds rate as those forecasts were revised. If viewed in the context of the Taylor (1993) Rule, discretionary departures from the settings prescribed by a Greenbook forecast-based version of the rule consistently presage business cycle turning points. Similarly, estimates from an interest rate rule with time-varying parameters imply that, around such turning points, the FOMC responds less vigorously to information contained in Greenbook forecasts about the changing state of the economy. These results suggest possible gains from closer adherence to a rule with constant parameters. Other statistical properties of Greenbook forecasts also point to an overlooked role for monetary aggregates, particularly Divisia monetary aggregates, in the Federal Reserve's forecasting process and subsequent monetary policy decisions made by the FOMC.
    Keywords: Greenbook forecasts, Taylor Rule, Time-varying parameters, Divisia monetary aggregates
    JEL: E31 E32 E37 E43 E47 E51 E52 E58 E65
    Date: 2018–07–01
  12. By: Raghavan, Mala (Tasmanian School of Business & Economics, University of Tasmania); Athanasopoulos, George (Monash University)
    Abstract: Using a parsimonious structural vector autoregressive moving average (SVARMA) model, we analyse the transmission of foreign and domestic shocks to a small open emerging economy under di erent policy regimes. Narrower con dence bands around the SVARMA responses compared to the SVAR responses, advocate the suitability of this framework for analysing the propagation of economic shocks over time. Malaysia is an interesting small open economy that has experienced an ongoing process of economic transition and development. The Malaysian government imposed exchange rate and capital control measures following the 1997 Asian nancial crisis. Historical and variance decompositions highlight Malaysia's high exposure to foreign shocks. The effects of these shocks change over time under di erent policy regimes. During the pegged exchange rate period, Malaysian monetary policymakers experienced some breathing space to focus on maintaining price and output stability. In the post-pegged period, Malaysia's exposure to foreign shocks increased and in recent times are largely driven by world commodity price and global activity shocks.
    Keywords: SVARMA models, Open Economy Macroeconomics, ASEAN, Shock transmissions
    JEL: C32 F41 E52
    Date: 2018

This nep-for issue is ©2018 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.