nep-for New Economics Papers
on Forecasting
Issue of 2013‒11‒29
eleven papers chosen by
Rob J Hyndman
Monash University

  1. Do high-frequency financial data help forecast oil prices? The MIDAS touch at work By Baumeister, Christiane; Guérin, Pierre; Kilian, Lutz
  2. Forecasting Business Investment in the Short Term Using Survey Data By Österholm, Pär
  3. Probability and Severity of Recessions By Rachidi Kotchoni; dalibor Stevanovic
  4. “Tourism demand forecasting with different neural networks models” By Oscar Claveria; Enric Monte; Salvador Torra
  5. “Forecasting Business surveys indicators: neural networks vs. time series models” By Oscar Claveria; Salvador Torra
  6. DSGE Model-Based Forecasting of Modeled and Non-Modeled Inflation Variables in South Africa By Rangan Gupta; Patrick T. kanda; Mampho P. Modise; Alessia Pacagnini
  7. Forecasting the NOK/USD Exchange Rate with Machine Learning Techniques By Theophilos Papadimitriou; Periklis Gogas; Vasilios Plakandaras
  8. Some thoughts on making long-term forecasts for the world economy By Fardoust, Shahrokh; Dhareshwar, Ashok
  9. Sparse canonical correlation analysis from a predictive point of view. By Wilms, Ines; Croux, Christophe
  10. Eliciting Private Information with Noise: The Case of Randomized Response By Andreas Blume; Ernest K. Lai; Wooyoung Lim
  11. Behavioural Asymmetries in the G7 Foreign Exchange Market By mamatzakis, e; Christodoulakis, G

  1. By: Baumeister, Christiane; Guérin, Pierre; Kilian, Lutz
    Abstract: The substantial variation in the real price of oil since 2003 has renewed interest in the question of how to forecast monthly and quarterly oil prices. There also has been increased interest in the link between financial markets and oil markets, including the question of whether financial market information helps forecast the real price of oil in physical markets. An obvious advantage of financial data in forecasting oil prices is their availability in real time on a daily or weekly basis. We investigate whether mixed-frequency models may be used to take advantage of these rich data sets. We show that, among a range of alternative high-frequency predictors, especially changes in U.S. crude oil inventories produce substantial and statistically significant real-time improvements in forecast accuracy. The preferred MIDAS model reduces the MSPE by as much as 16 percent compared with the no-change forecast and has statistically significant directional accuracy as high as 82 percent. This MIDAS forecast also is more accurate than a mixed-frequency realtime VAR forecast, but not systematically more accurate than the corresponding forecast based on monthly inventories. We conclude that typically not much is lost by ignoring high-frequency financial data in forecasting the monthly real price of oil. --
    Keywords: Mixed frequency,Real-time data,Oil price,Forecasts
    JEL: C53 G14 Q43
    Date: 2013
  2. By: Österholm, Pär (National Institute of Economic Research)
    Abstract: In this paper, forecasting models for Swedish business investment growth which make use of data from Sweden’s most important business survey – the Economic Tendency Survey – are evaluated. An out-of-sample forecast exercise using nine years of quarterly real-time data is conducted. The results suggest that the survey data have informational value that can be used to improve forecasts. Perhaps not surprisingly, the time series with the highest predictive power for business investment growth tend to be based on data for the investment goods industry. Forecasts based on a simple arithmetic mean of individual model forecasts do well in the evaluation and should accordingly be useful when forecasting Swedish business investment in practice.
    Keywords: Out-of-sample forecasts; Real-time data
    JEL: E22 E27
    Date: 2013–11–19
  3. By: Rachidi Kotchoni; dalibor Stevanovic
    Abstract: This paper advances beyond the prediction of the probability of a recession by also considering its severity in terms of output loss and duration. First, Probit models are used to estimate the probability of a recession at period t + h from the information available at period t. Next, a Vector Autoregression (VAR) augmented with diffusion indices and an inverse Mills ratio (IMR) is fitted to selected measures of real economic activity. The latter model is used to generate two forecasts: an average forecast, and a forecast under pessimistic assumption that a recession occurs at the forecast horizon. The severity of recessions is then predicted as the gap between these two forecasts. Finally, a zero-inflated Poisson model is fitted to historical durations of recessions. Our empirical results suggest that U.S. recessions are fairly predictable, both in terms of occurrence and severity. Out-of-sample experiments suggest that the inclusion of the IMR in the VAR model significantly improves its forecasting performance.
    Keywords: Duration of recession, Forecasting Real Activity, Probability of Recessions, Probit, Vector Autoregression, Zero Inflated Poisson
    JEL: C3 C5 C35 E27 E37
    Date: 2013
  4. By: Oscar Claveria (Faculty of Economics, University of Barcelona); Enric Monte (Department of Signal Theory and Communications, Polytechnic University of Catalunya (UPC)); Salvador Torra (Faculty of Economics, University of Barcelona)
    Abstract: This paper aims to compare the performance of different Artificial Neural Networks techniques for tourist demand forecasting. We test the forecasting accuracy of three different types of architectures: a multi-layer perceptron, a radial basis function and an Elman network. We also evaluate the effect of the memory by repeating the experiment assuming different topologies regarding the number of lags introduced. We used tourist arrivals from all the different countries of origin to Catalonia from 2001 to 2012. We find that multi-layer perceptron and radial basis function models outperform Elman networks, being the radial basis function architecture the one providing the best forecasts when no additional lags are incorporated. These results indicate the potential existence of instabilities when using dynamic networks for forecasting purposes. We also find that for higher memories, the forecasting performance obtained for longer horizons improves, suggesting the importance of increasing the dimensionality for long term forecasting.
    Keywords: tourism demand; forecasting; artificial neural networks; multi-layer perceptron; radial basis function; Elman networks; Catalonia. JEL classification: L83; C53; C45; R11
    Date: 2013–11
  5. By: Oscar Claveria (Faculty of Economics, University of Barcelona); Salvador Torra (Faculty of Economics, University of Barcelona)
    Abstract: The objective of this paper is to compare different forecasting methods for the short run forecasting of Business Survey Indicators. We compare the forecasting accuracy of Artificial Neural Networks (ANN) vs. three different time series models: autoregressions (AR), autoregressive integrated moving average (ARIMA) and self-exciting threshold autoregressions (SETAR). We consider all the indicators of the question related to a country’s general situation regarding overall economy, capital expenditures and private consumption (present judgement, compared to same time last year, expected situation by the end of the next six months) of the World Economic Survey (WES) carried out by the Ifo Institute for Economic Research in co-operation with the International Chamber of Commerce. The forecast competition is undertaken for fourteen countries of the European Union. The main results of the forecast competition are offered for raw data for the period ranging from 1989 to 2008, using the last eight quarters for comparing the forecasting accuracy of the different techniques. ANN and ARIMA models outperform SETAR and AR models. Enlarging the observed time series of Business Survey Indicators is of upmost importance in order of assessing the implications of the current situation and its use as input in quantitative forecast models.
    Keywords: Business surveys; Forecasting; Time series models; Nonlinear models; Neural networks.
    Date: 2013–11
  6. By: Rangan Gupta (Department of Economics, University of Pretoria); Patrick T. kanda (Department of Economics, University of Pretoria); Mampho P. Modise (Department of Economics, University of Pretoria); Alessia Pacagnini (Dipartimento di Economia, Metodi Quantitativi e Strategie d'Impresa (DEMS), Facoltà di Economia, Università degli Studi di Milano-Bicocca)
    Abstract: Inflation forecasts are a key ingredient for monetary policymaking - especially in an inflation targeting country such as South Africa. Generally, a typical Dynamic Stochastic General Equilibrium (DSGE) only includes a core set of variables. As such, other variables,e.g. such as alternative measures of inflation that might be of interest to policymakers, do not feature in the model. Given this, we implement a closed-economy New Keynesian DSGE model-based procedure which includes variables that do not explicitly appear in the model. We estimate such a model using an in-sample covering 1971Q2 to 1999Q4, and generate recursive forecasts over 2000Q1-2011Q4. The hybrid DSGE performs extremely well in forecasting inflation variables (both core and non-modeled) in comparison with forecasts reported by other models, such as the AR(1).
    Keywords: DSGE model, in ation, core variables, non-core variables
    JEL: C11 C32 C53 E27 E47
    Date: 2013–11
  7. By: Theophilos Papadimitriou (Department of Economics, University Campus Komotini, Democritus University of Thrace, Greece); Periklis Gogas (Department of Economics, University Campus Komotini, Democritus University of Thrace, Greece); Vasilios Plakandaras (Department of Economics, University Campus Komotini, Democritus University of Thrace, Greece)
    Abstract: In this paper, we approximate the empirical findings of Papadamou and Markopoulos (2012) on the NOK/USD exchange rate under a Machine Learning (ML) framework. By applying Support Vector Regression (SVR) on a general monetary exchange rate model and a Dynamic Evolving Neuro-Fuzzy Inference System (DENFIS) to extract model structure, we test for the validity of popular monetary exchange rate models. We reach to mixed results since the coefficient sign of interest rate differential is in favor only with the model proposed by Bilson (1978), while the inflation rate differential coefficient sign is approximated by the model of Frankel (1979). By adopting various inflation expectation estimates, our SVR model fits actual data with a small Mean Absolute Percentage Error when an autoregressive approach excluding energy prices is adopted for inflation expectation. Overall, our empirical findings conclude that for a small open petroleum producing country such as Norway, fundamentals possess significant forecasting ability when used in exchange rate forecasting.
    Keywords: International Financial Markets, Foreign Exchange, Support Vector Regression, Monetary exchange rate models
    JEL: G15 F30 F31
    Date: 2013–11
  8. By: Fardoust, Shahrokh; Dhareshwar, Ashok
    Abstract: Countries and international organizations working on longer-range development issues depend on long-term quantitative projections and scenario analysis. Such forecasting has become increasingly challenging, thanks to the rapid pace of globalization, technological progress, the interplay among them, and enhanced connectivity among people. As a result, seemingly isolated events can quickly lead to wide-ranging and lasting regional or even global consequences. This paper examines the problem of long-term economic forecasting in the face of increased complexity and uncertainty. With the benefit of hindsight, it scrutinizes past long-term qualitative and quantitative projections for the 1990s in order to draw lessons on how an institution can and should conduct long-term forecasting and policy analysis. The main conclusions are that policy makers and researchers across the world urgently need to see the big picture if they are to deal with the specific challenges and opportunities they will face over the long term as economies and global linkages undergo major structural changes under conditions of considerable uncertainty and volatility. Global institutions need to have strong research programs that work in close collaboration with other international organizations, academic centers, and independent experts on important long-term development issues ("blue sky"issues) and megatrends. These institutions need to build on their comparative strengths and form teams of in-house researchers and global experts who work on state-of-the-art models related to globalization, technological progress and innovations, climate change, demographic shifts, population, and labor force quality and their policy implications at both the global and country levels. Researchers should be encouraged to consider how global challenges such as financial crises, climate change, and infectious diseases can lead to breaks in economic trends and regime change and how such breaks affect economic activity. Alternative scenarios need to be created that incorporate the views of contrarian forecasters, including forecasts of possible shocks.
    Keywords: Environmental Economics&Policies,Economic Theory&Research,Emerging Markets,Currencies and Exchange Rates,Banks&Banking Reform
    Date: 2013–11–01
  9. By: Wilms, Ines; Croux, Christophe
    Abstract: Canonical correlation analysis (CCA) describes the associations between two sets of variables by maximizing the correlation between linear combinations of the variables in each data set. However, in high-dimensional settings where the number of variables exceeds the sample size or when the variables are highly correlated, traditional CCA is no longer appropriate. This paper proposes a method for sparse CCA. Sparse estimation produces linear combinations of only a subset of variables from each data set, thereby increasing the interpretability of the canonical variates. We consider the CCA problem from a predictive point of view and recast it into a multivariate regression framework. By combining a multivariate alternating regression approach together with a lasso penalty, we induce sparsity in the canonical vectors. We compare the performance with other sparse CCA techniques in dierent simulation settings and illustrate its usefulness on a genomic data set.
    Keywords: Canonical correlation analysis; Genomic data; Lasso; Multivariate regression; Sparsity;
    Date: 2013–11
  10. By: Andreas Blume; Ernest K. Lai; Wooyoung Lim
    Abstract: The paper formalizes Warner's (1965) randomized response technique (RRT) as a game and implements it experimentally, thus linking game theoretic approaches to randomness in communication with survey practice in the field and a novel implementation in the lab. As predicted by our model and in line with Warner, the frequency of truthful responses is significantly higher with randomization than without. The model predicts that randomization weakly improves information elicitation, as measured in terms of mutual information, although, surprisingly, not always by RRT inducing truth-telling. Contrary to this prediction, randomization significantly reduces the elicited information in our experiment.
    Keywords: Randomized Response, Lying Aversion, Stigmatization Aversion, Mutual Information, Laboratory Experiments
    JEL: C72 C92 D82 D83
    Date: 2013–07
  11. By: mamatzakis, e; Christodoulakis, G
    Abstract: This paper examines the exchange rate disconnect puzzle of Obstfeld and Rogoff, (2000) from a behavioural perspective. It provides evidence on the existence of substantial asymmetries in the underlying loss preferences for the difference between the spot and forward nominal exchange rates between the G7 countries for one-week and four-week forecast horizons. We further perform forecast breakdown tests in forward markets during the Greek and the Portuguese sovereign debt crisis, and then re-estimate the loss preferences showing a mean-reverting transition from optimism to pessimism and vice versa. Finally, we attribute the evolution of preferences to economic fundamentals and risk indexes and find that together with significant endogenous dynamics, variables such as growth and deficit differentials, interest rate and legal risk assert some significant impact on asymmetry. This new set of information suggests that the puzzle could have its roots on an underlying asymmetric loss function that reflects variability in preferences over exchange rate movements due to a variety of episodes in economic fundamentals.
    Keywords: Spot-forward exchange rates, Asymmetric preferences, Forecast breakdown, GMM estimation.
    JEL: F31 F47
    Date: 2013–11–19

This nep-for issue is ©2013 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.