nep-for New Economics Papers
on Forecasting
Issue of 2015‒01‒14
eleven papers chosen by
Rob J Hyndman
Monash University

  1. The Fluke Of Stochastic Volatility Versus Garch Inevitability : Which Model Creates Better Forecasts? By Valeria V. Lakshina
  2. Using Entropic Tilting to Combine BVAR Forecasts with External Nowcasts By Krueger, Fabian; Clark, Todd E.; Ravazzolo, Francesco
  3. Optimal forecasts from Markov switching models By Tom Boot; Andreas Pick
  4. A multi-country approach to forecasting output growth using PMIs By Chudik, Alexander; Grossman, Valerie; Pesaran, M. Hashem
  5. A large Bayesian vector autoregression model for Russia By Deryugina , Elena; Ponomarenko , Alexey
  6. An Application of a Short Memory Model With Random Level Shifts to the Volatility of Latin American Stock Market Returns By Gabriel Rodriguez; Roxana Tramontana
  7. Forecast combinations in a DSGE-VAR lab By Costantini, Mauro; Gunter, Ulrich; Kunst, Robert M.
  8. Do Macroeconomic Shocks Affect Intuitive Inflation Forecasting? An Experimental Investigation By Marvin Deversi
  9. Nowcasting Tourist Arrivals to Prague: Google Econometrics By Zeynalov, Ayaz
  10. Predicting US Recessions: Does a Wishful Bias Exist? By Sergey V. Smirnov
  11. Forecasting day ahead electricity spot prices: The impact of the EXAA to other European electricity markets By Florian Ziel; Rick Steinert; Sven Husmann

  1. By: Valeria V. Lakshina (National Research University Higher School)
    Abstract: The paper proposes the thorough investigation of the in-sample and out-of-sample performance of four GARCH and two stochastic volatility models, which were estimated based on Russian financial data. The data includes Aeroflot and Gazprom’s stock prices, and the rouble against the US dollar exchange rates. In our analysis, we use the probability integral transform for the in-sample comparison, and a Mincer-Zarnowitz regression, along with classical forecast performance measures, for the out-of-sample comparison. Studying both the explanatory and the forecasting power of the models analyzed, we came to the conclusion that stochastic volatility models perform equally or in some cases better than GARCH models.
    Keywords: GARCH, stochastic volatility, markov switching multifractal, forecast performance.
    JEL: C01 C58 C51 G17
    Date: 2014
  2. By: Krueger, Fabian (Heidelburg Institute for Theoretical Studies); Clark, Todd E. (Federal Reserve Bank of Cleveland); Ravazzolo, Francesco (Norges Bank and the BI Norwegian Business School)
    Abstract: This paper shows entropic tilting to be a flexible and powerful tool for combining medium-term forecasts from BVARs with short-term forecasts from other sources (nowcasts from either surveys or other models). Tilting systematically improves the accuracy of both point and density forecasts, and tilting the BVAR forecasts based on nowcast means and variances yields slightly greater gains in density accuracy than does just tilting based on the nowcast means. Hence entropic tilting can offer—more so for persistent variables than not-persistent variables—some benefits for accurately estimating the uncertainty of multi-step forecasts that incorporate nowcast information.
    Keywords: Forecasting; Prediction; Bayesian Analysis
    JEL: C11 C53 E17
    Date: 2015–01–07
  3. By: Tom Boot; Andreas Pick
    Abstract: We derive optimal weights for Markov switching models by weighting observations such that forecasts are optimal in the MSFE sense. We provide analytic expressions of the weights conditional on the Markov states and conditional on state probabilities. This allows us to study the effect of uncertainty around states on forecasts. It emerges that, even in large samples, forecasting performance increases substantially when the construction of optimal weights takes uncertainty around states into account. Performance of the optimal weights is shown through simulations and an application to US GNP, where using optimal weights leads to significant reductions in MSFE.
    Keywords: Markov switching models; forecasting; optimal weights; GNP forecasting
    JEL: C25 C53 E37
    Date: 2014–12
  4. By: Chudik, Alexander (Federal Reserve Bank of Dallas); Grossman, Valerie (Federal Reserve Bank of Dallas); Pesaran, M. Hashem (University of Southern California)
    Abstract: This paper derives new theoretical results for forecasting with Global VAR (GVAR) models. It is shown that the presence of a strong unobserved common factor can lead to an undetermined GVAR model. To solve this problem, we propose augmenting the GVAR with additional proxy equations for the strong factors and establish conditions under which forecasts from the augmented GVAR model (AugGVAR) uniformly converge in probability (as the panel dimensions N,T→ ∞ such that N/T→κ for some 0
    JEL: C53 E37
    Date: 2014–11–01
  5. By: Deryugina , Elena (BOFIT); Ponomarenko , Alexey (BOFIT)
    Abstract: We apply an econometric approach developed specifically to address the ‘curse of dimensionality’ in Russian data and estimate a Bayesian vector autoregression model comprising 14 major domestic real, price and monetary macroeconomic indicators as well as external sector variables. We conduct several types of exercise to validate our model: impulse response analysis, recursive forecasting and counter factual simulation. Our results demonstrate that the employed methodology is highly appropriate for economic modelling in Russia. We also show that post-crisis real sector developments in Russia could be accurately forecast if conditioned on the oil price and EU GDP (but not if conditioned on the oil price alone).
    Keywords: Bayesian vector autoregression; forecasting; Russia
    JEL: C32 E32 E44 E47
    Date: 2014–12–03
  6. By: Gabriel Rodriguez (Departamento de Economía - Pontificia Universidad Católica del Perú); Roxana Tramontana (Departamento de Economía - Pontificia Universidad Católica del Perú)
    Abstract: Empirical research indicates that the volatility of stock return time series have long memory. However, it has been demonstrated that short memory processes contaminated with random level shifts can often be confused as being long memory. Often this feature is referred to as spurious long memory. This paper represents an empirical study of the random level shift (RLS) model using the approach of Lu and Perron (2010) and Li and Perron (2013) for the volatility of daily stocks returns data for Öve Latin American countries. The RLS model consists of the sum of a short term memory component and a level shift component, where the level shift component is governed by a Bernoulli process with a shift probability . The estimation results suggest that the level shifts in the volatility of daily stocks returns data are infrequent but once they are taken into account, the long memory characteristic and the GARCH e§ects disappear. An out-of-sample forecasting exercise is also provided. JEL Classification-JEL: C22
    Keywords: Returns, Volatility, Long Memory, Random Level Shifts, Kalman Filter, Forecasting, Latin America
    Date: 2014
  7. By: Costantini, Mauro (Department of Economics and Finance, Brunel University); Gunter, Ulrich (Department of Tourism and Service Management, MODUL University Vienna); Kunst, Robert M. (Department of Economics and Finance, Institute for Advanced Studies, Vienna and Department of Economics, University of Vienna)
    Abstract: We explore the benefits of forecast combinations based on forecast-encompassing tests compared to simple averages and to Bates-Granger combinations. We also consider a new combination method that fuses test-based and Bates-Granger weighting. For a realistic simulation design, we generate multivariate time-series samples from a macroeconomic DSGE-VAR model. Results generally support Bates-Granger over uniform weighting, whereas benefits of test-based weights depend on the sample size and on the prediction horizon. In a corresponding application to real-world data, simple averaging performs best. Uniform averages may be the weighting scheme that is most robust to empirically observed irregularities.
    Keywords: Combining forecasts, encompassing tests, model selection, time series, DSGE-VAR model
    Date: 2014–12
  8. By: Marvin Deversi
    Abstract: In an experimental setting impulse-response behaviour in intuitive inflation forecasting is analysed. Participants were asked to forecast future values of inflation for a fictitious economy after receiving charts and lists of past values of inflation and output gap. Thirty periods were forecasted stepwise and feedback on performance was provided after each period. In a between subjects design, participants experienced a negative or positive supply shock. The results suggest that participants barely report rational forecasts. Instead, simple backward-looking rules describe stated forecast series. Forecasting is heterogeneous across agents and over time. Before the shock, most participants can be described by natural expectations. Due to the shocks 69% of participants are found to switch their forecasting rule. After the negative supply shock, subjects increase efficiency of forecasts. But, after a positive supply shock efficiency drops down to zero; this is evidence for a negativity bias. As a main result, macroeconomic shocks do alter the way experimental participants form intuitive inflation forecasts, however, to what extent depends on the shocks’ characteristics.
    Keywords: Macroeconomic experiment; inflation expectations; intuitive forecasting; shocks; heterogeneity
    JEL: C91 D84
    Date: 2014–12
  9. By: Zeynalov, Ayaz
    Abstract: It is expected that what people are searching for today is predictive of what they have done recently or will do in the near future. This research will analyze the eligibility of Google search data to nowcast tourist arrivals to Prague. The present research will report whether Google data is potentially useful in nowcasting or short-term forecasting using by Support Vector Regressions (SVRs), which maps data to a higher dimensional space and employs a kernel function.
    Keywords: Google trends, nowcasting, tourism forecasting
    JEL: C53 E17 L83
    Date: 2014
  10. By: Sergey V. Smirnov (National Research University Higher School of Economics)
    Abstract: There is evidence in the economic literature that near cyclical peaks an optimistic bias exists in private expert forecasts of real GDP growth rates. Other evidence concerns differences in the accuracy of GDP forecasts made during expansions and those made during contractions. It has also been hypothesized that a wishful bias may hamper the ability to recognize the beginning of a recession in real-time. We tested consensus forecasts of quarterly GDP growth rates taken from SPFs conducted by PhilFed and found that they may be seen as unbiased only for time horizons j=0,1,2; for greater horizons they are over-optimistic. This over-optimism may also be observed for (j=1, 2) for forecasts made at peaks (at these moments the consensus usually points only to a slowdown of the economy but not to a contraction). Lastly, over-optimism may be observed for nowcasts (j=0) during cyclical contractions, including the first two quarters of a recession (in these cases the reality is usually worse than expected). Taken together, all these facts mean that some aversion to predicting US recessions exists. There are two possible reasons for this: a) experts rely too heavily on extrapolations (then changes in medium-long tendencies would be missed in real time); b) there is a wishful bias in forecasts against predicting recessions (this reluctance may be rooted in psychological factors). We give some arguments in favor of the thesis that the second factor is more important.
    Keywords: Business cycles, Turning points, Recessions, Biased forecasts, SPF
    JEL: E32 E37
    Date: 2014
  11. By: Florian Ziel; Rick Steinert; Sven Husmann
    Abstract: In our paper we analyze the relationship between the day-ahead electricity price of the Energy Exchange Austria (EXAA) and other day-ahead electricity prices in Europe. We focus on markets, which settle their prices after the EXAA, which enables traders to include the EXAA price into their calculations. For each market we employ econometric models to incorporate the EXAA price and compare them with their counterparts without the price of the Austrian exchange. By employing a forecasting study, we find that electricity price models can be improved when EXAA prices are considered.
    Date: 2015–01

This nep-for issue is ©2015 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.