nep-for New Economics Papers
on Forecasting
Issue of 2017‒09‒03
six papers chosen by
Rob J Hyndman
Monash University

  1. Predicting US CPI-Inflation in the presence of asymmetries, persistence, endogeneity, and conditional heteroscedasticity By Afees A. Salisu; Kazeem Isah
  2. A multi-factor predictive model for oil-US stock nexus with persistence, endogeneity and conditional heteroscedasticity effects By Afees A. Salisu; Raymond Swaray; Tirimisyu F. Oloko
  3. Do Phillips Curves Conditionally Help to Forecast Inflation? By Dotsey, Michael; Fujita, Shigeru; Stark, Tom
  4. Identifying Exchange Rate Common Factors By Ryan Greenaway-McGrevy; Donggyu Sul; Nelson Mark; Jyh-Lin Wu
  5. Theories, techniques and the formation of German business cycle forecasts: Evidence from a survey among professional forecasters By Jörg Döpke; Ulrich Fritsche; Gabi Waldhof
  6. Improving the Predictive ability of oil for inflation: An ADL-MIDAS Approach. By Afees A. Salisu; Ahaemefula Ephraim Ogbonna

  1. By: Afees A. Salisu (Centre for Econometric and Allied Research, University of Ibadan); Kazeem Isah (Centre for Econometric and Allied Research, University of Ibadan)
    Abstract: In this paper, we construct a multi-predictor framework for US inflation by augmenting the traditional Phillips curve-based inflation model with symmetric and asymmetric oil price changes. We show that the underlying predictors of US inflation exhibit persistence, endogeneity and conditional heteroscedasticity effects which have implications on forecast performance. Thus, we employ the Westerlund and Narayan (WN hereafter) (2014) estimator which allows for these effects in the predictive model. Also, we follow the linear multi-predictor set-up by Makin et al. (2014) which is an extension of the bivariate predictive model of WN (2014). Thereafter, we extend the former in order to construct a nonlinear multi-predictor model that allows for asymmetries based on Shin et al. (2014) approach. Using historical quarterly data for relevant variables ranging from 1957 to 2017, we demonstrate that US inflation is better modelled with the proposed multi-predictor model suggesting the significance of oil price in the predictive model for US inflation. In addition, we find that the US inflation forecast is episodic and asymmetric. Among the competing multi-predictor variants, the positive oil price-based variant outperforms all other variants both for in-sample and out-of-sample forecasts. The proposed model also outperforms the autoregressive process for a longer out-of-sample period. Our results are robust to different measures of inflation, multiple in-sample periods and forecast horizon.
    Keywords: OECD, Phillips curve, Asymmetries, Inflation forecasts, Forecast evaluation
    JEL: C53 E31 E37
    Date: 2017–08
  2. By: Afees A. Salisu (Centre for Econometric and Allied Research, University of Ibadan); Raymond Swaray (Economics Subject Group, University of Hull Business, University of Hull, Cottingham Road, UK); Tirimisyu F. Oloko (Centre for Econometric and Allied Research, University of Ibadan)
    Abstract: In this study, we extend the single-predictor model for US stock market developed by Narayan and Gupta (2014) to capture more important predictors of the market. Our analyses are conducted in three distinct ways. First, we test whether oil price will produce better forecast accuracy in the multiple-factor model than in the single-factor model. Secondly, we also test the plausibility of making generalization about the predictive model for oil-US stocks on the basis of large cap stocks. Thirdly, we employ the recently developed Feasible Quasi Generalized Least Squares (FQGLS) estimator by Westerlund and Narayan (2014) in order to capture the inherent persistence, endogeneity and heteroscedasticity effects in the predictors. Our results reveal that oil price renders better forecast performance in the multiple-factor predictive model than in the single-factor variants for both in-sample and out-of-sample forecasts. Also, we find that generalizing the predictability of oil-US stock market with large cap may lead to misleading inferences. In addition, it may be necessary to pre-test the predictors for persistence, endogeneity and conditional heteroscedasticity particularly when modeling with high frequency series. Our results are robust to different forecast measures and forecast horizons.
    Keywords: WTI Oil price; US large cap; US Small cap; Inflation; Output, Forecast evaluation
    JEL: G11 Q43
    Date: 2017–08
  3. By: Dotsey, Michael (Federal Reserve Bank of Philadelphia); Fujita, Shigeru (Federal Reserve Bank of Philadelphia); Stark, Tom (Federal Reserve Bank of Philadelphia)
    Abstract: This paper reexamines the forecasting ability of Phillips curves from both an unconditional and conditional perspective by applying the method developed by Giacomini and White (2006). We find that forecasts from our Phillips curve models tend to be unconditionally inferior to those from our univariate forecasting models. Significantly, we also find conditional inferiority, with some exceptions. When we do find improvement, it is asymmetric - Phillips curve forecasts tend to be more accurate when the economy is weak and less accurate when the economy is strong. Any improvement we find, however, vanished over the post-1984 period.
    Keywords: Phillips curve; unemployment gap; conditional predictive ability
    JEL: C53 E37
    Date: 2017–08–21
  4. By: Ryan Greenaway-McGrevy; Donggyu Sul; Nelson Mark; Jyh-Lin Wu
    Abstract: Using recently developed model selection procedures, we determine that exchange rate returns are driven by a two-factor model. We identify them as a dollar factor and a euro factor. Exchange rates are thus driven by global, US, and Euro-zone stochastic discount factors. The identified factors can also be given a risk-based interpretation. Identification motivates multilateral models for bilateral exchange rates. Out-of-sample forecast accuracy of empirically identified multilateral models dominate the random walk and a bilateral purchasing power parity fundamentals prediction model. 24-month ahead forecast accuracy of the multilateral model dominates those of a principal components forecasting model.
    JEL: F31 F37
    Date: 2017–08
  5. By: Jörg Döpke (University of Applied Sciences Merseburg (Hochschule Merseburg)); Ulrich Fritsche (Universität Hamburg); Gabi Waldhof (Martin-Luther-Universität Halle-Wittenberg)
    Abstract: The paper reports results of a survey among active forecasters of the German business cycle. Relying on 82 respondents from 37 different institutions, we investigate what models and theories forecasters subscribe to and find that they are pronounced conservative in the sense, that they overwhelmingly rely on methods and theories that have been well-established for a long time, while more recent approaches are relatively unimportant for the practice of business cycle forecasting. DSGE models are mostly used in public institutions. In line with findings in the literature there are tendencies of “leaning towards consensus” (especially for public institutions) and “sticky adjustment of forecasts” with regard to new information. We find little evidence that the behaviour of forecasters has changed fundamentally since the Great Recession but there are signs that forecast errors are evaluated more carefully. Also, a stable relationship between preferred theories and methods and forecast accuracy cannot be established.
    Keywords: Forecast error evaluation, questionnaire, survey, business cycle forecast, professional forecaster
    JEL: E32 E37 C83
    Date: 2017–08
  6. By: Afees A. Salisu (Centre for Econometric and Allied Research, University of Ibadan); Ahaemefula Ephraim Ogbonna (Centre for Econometric and Allied Research, University of Ibadan)
    Abstract: This paper attempts to improve the predictive ability of oil for inflation by incorporating mixed data sampling regression model into the autoregressive distributed lag model. The efficiency of the conventionally used models, which are based on same frequency of variables, is challenged on the basis of the concealed information in low frequency series. Using data covering OECD countries, we find that the ADL-MIDAS seems to outperform all the other competing models, a feat attributable to the integration of more information from a higher frequency oil price series in the forecast of a low frequency inflation series. In addition, including oil price in inflation model produces more accurate results than the model that excludes it.
    Keywords: OECD countries, ADL-MIDAS, Inflation forecasts, Forecast evaluation
    JEL: C53 E31 E37
    Date: 2017–08

This nep-for issue is ©2017 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.