|
on Forecasting |
| By: | Clements, Michael P (University of Warwick) |
| Abstract: | We consider whether survey respondents’probability distributions, reported as histograms, provide reliable and coherent point predictions, when viewed through the lens of a Bayesian learning model, and whether they are well calibrated more generally. We argue that a role remains for eliciting directly-reported point predictions in surveys of professional forecasters. Key words: probability distribution forecasts ; point forecasts ; Bayesian learning JEL classification: C53 |
| Date: | 2012 |
| URL: | http://d.repec.org/n?u=RePEc:wrk:warwec:976&r=for |
| By: | Ulrich Fritsche (Department for Socioeconomics, Department for Economics, University of Hamburg); Christian Pierdzioch (Helmut-Schmidt-University, Department of Economics); Jan-Christoph Ruelke (WHU – Otto Beisheim School of Management); Georg Stadtmann (University of Southern Denmark, Department of Business and Economics, and European-University Viadrina) |
| Abstract: | Based on the approach advanced by Elliott et al. (Rev. Ec. Studies. 72, 1197-1125,2005), we analyzed whether the loss function of a sample of exchange rate forecasters is asymmetric in the forecast error. Using forecasts of the euro/dollar exchange rate, we found that the shape of the loss function varies across forecasters. Our empirical results suggest that it is important to account for the heterogeneity of exchange rate forecasts at the microeconomic level of individual forecasters when one seeks to analyze whether forecasters form exchange rate forecasts under an asymmetric loss function. |
| Keywords: | Exchange rate, Forecasting, Loss function |
| JEL: | F31 D84 |
| Date: | 2012–02 |
| URL: | http://d.repec.org/n?u=RePEc:hep:macppr:201201&r=for |
| By: | Antipin, Jan-Erik (National Institute of Economic Research); Boumediene, Farid Jimmy (Ministry of Finance); Österholm, Pär (Sveriges Riksbank) |
| Abstract: | This paper assesses the usefulness of constant gain least squares when forecasting inflation. An out-of-sample forecast exercise is conducted, in which univariate autoregressive models for inflation in Australia, Swe-den, the United Kingdom and the United States are used. The results suggest that it is possible to improve the forecast accuracy by employing constant gain least squares instead of ordinary least squares. In particular, when using a gain of 0.05, constant gain least squares generally outper-forms the corresponding autoregressive model estimated with ordinary least squares. In fact, at longer forecast horizons, the root mean square forecast error is reliably lowered for all four countries and for all lag lengths considered in the study. |
| Keywords: | Out-of-sample forecasts; Inflation |
| JEL: | E31 E37 |
| Date: | 2012–02–01 |
| URL: | http://d.repec.org/n?u=RePEc:hhs:nierwp:0126&r=for |
| By: | Ubilava, David; Helmers, C Gustav |
| Abstract: | This study examines the benets of nonlinear time series modelling to improve forecast accuracy of the El Nino Southern Oscillation (ENSO) phenomenon. The paper adopts a smooth transition autoregressive (STAR) modelling framework to assess the potentially regime-dependent dynamics of sea surface temperature anomaly. The results reveal STAR-type nonlinearities in ENSO dynamics, resulting in superior out-of-sample forecast performance of STAR over the linear autoregressive models. The advantage of nonlinear models is especially apparent in the short- and intermediate-term forecasts. These results are of interest to researchers and policy makers in the elds of climate dynamics, agricultural production, and environmental management. |
| Keywords: | El Nino Southern Oscillation; Out-of-Sample Forecasting; Smooth Transition Autoregression |
| JEL: | C53 C22 Q54 |
| Date: | 2012–01 |
| URL: | http://d.repec.org/n?u=RePEc:pra:mprapa:36890&r=for |
| By: | Ulrich Fritsche (Department for Socioeconomics, Department for Economics, University of Hamburg); Christian Pierdzioch (Helmut-Schmidt-University, Department of Economics); Jan-Christoph Ruelke (WHU – Otto Beisheim School of Management); Georg Stadtmann (University of Southern Denmark, Department of Business and Economics, and European-University Viadrina) |
| Abstract: | Using forecasts of the Brazilian real and the Mexican peso, we analyze the shape of the loss function of exchange-rate forecasters and the rationality of their forecasts. We find a substantial degree of cross-sectional heterogeneity with respect to the shape of the loss function. While some forecasters seem to forecasts under an asymmetric loss function, symmetry of the loss function cannot be rejected for other forecasters. An asymmetric loss function does not necessarily make survey data of exchange-rate forecasts look rational, and the loss function seems to depend not only on the forecast error. |
| Keywords: | Exchange rate, Forecasting, Loss function |
| JEL: | F31 D84 |
| Date: | 2012–02 |
| URL: | http://d.repec.org/n?u=RePEc:hep:macppr:201202&r=for |
| By: | Simón Sosvilla-Rivero (Universidad Complutense de Madrid); Maria del Carmen Ramos-Herrera (Universidad Complutense de Madrid) |
| Abstract: | We examine the predictive ability and consistency properties of exchange rate expectations for the dollar/euro using a survey conducted in Spain by PwC among a panel of experts and entrepreneurs. Our results suggest that the PwC panel have some forecasting ability for time horizons from 3 to 9 months, although only for the 3-month ahead expectations we obtain marginal evidence of unbiasedness and efficiency in the forecasts. As for the consistency properties of the exchange rate expectations formation process, we find that survey participants form stabilising expectations in the short-run and destabilising expectations in the long- run and that the expectation formation process is closer to fundamentalists than chartists. |
| Keywords: | Exchange rates, Forecasting; Expectations; Panel data; Econometric models |
| JEL: | F31 D84 C33 |
| Date: | 2012–02 |
| URL: | http://d.repec.org/n?u=RePEc:aee:wpaper:1202&r=for |
| By: | Michael Wolf; Dan Wunderli |
| Abstract: | Many economic and financial applications require the forecast of a random variable of interest over several periods into the future. The sequence of individual forecasts, one period at a time, is called a path-forecast, where the term path refers to the sequence of individual future realizations of the random variable. The problem of constructing a corresponding joint prediction region has been rather neglected in the literature so far: such a region is supposed to contain the entire future path with a prespecified probability. We develop bootstrap methods to construct joint prediction regions. The resulting regions are proven to be asymptotically consistent under a mild high-level assumption. We compare the finite-sample performance of our joint prediction regions to some previous proposals via Monte Carlo simulations. An empirical application to a real data set is also provided. |
| Keywords: | Generalized error rates, path-forecast, simultaneous prediction intervals |
| JEL: | C14 C32 C53 |
| Date: | 2012–02 |
| URL: | http://d.repec.org/n?u=RePEc:zur:econwp:064&r=for |
| By: | Claudia Godbout; Marco J. Lombardi |
| Abstract: | While the usefulness of factor models has been acknowledged over recent years, little attention has been devoted to the forecasting power of these models for the Japanese economy. In this paper, we aim at assessing the relative performance of factor models over different samples, including the recent financial crisis. To do so, we construct factor models to forecast Japanese GDP and its subcomponents, using 38 data series (including daily, monthly and quarterly variables) over the period 1991 to 2010. Overall, we find that factor models perform well at tracking GDP movements and anticipating turning points. For most of the components, we report that factor models yield lower forecasting errors than a simple AR process or an indicator model based on Purchasing Managers’ Indicators (PMIs). In line with previous studies, we conclude that the largest improvements in terms of forecasting accuracy are found for more volatile periods, such as the recent financial crisis. However, unlike previous studies, we do not find evident links between the volatility of the components and the relative advantage of using factor models. Finally, we show that adding the PMI index as an independent explanatory variable improves the forecasting properties of the factor models. |
| Keywords: | Econometric and statistical methods; International topics |
| JEL: | C50 C53 E37 E47 |
| Date: | 2012 |
| URL: | http://d.repec.org/n?u=RePEc:bca:bocawp:12-7&r=for |
| By: | Prieto Rodríguez, Juan; Ateca Amestoy, Victoria María |
| Abstract: | A first version of this paper was presented at the University of Catania, 2011 and at the fifth European Workshop on Applied Cultural Economics in Dublin, 2011. |
| Keywords: | forecasting, count data, prediction intervals, Brier scores, bootstrapping, arts |
| JEL: | Z11 D12 |
| Date: | 2012–01 |
| URL: | http://d.repec.org/n?u=RePEc:ehu:dfaeii:201201&r=for |
| By: | Maheu, John; Song, Yong |
| Abstract: | This paper develops an efficient approach to model and forecast time-series data with an unknown number of change-points. Using a conjugate prior and conditional on time-invariant parameters, the predictive density and the posterior distribution of the change-points have closed forms. The conjugate prior is further modeled as hierarchical to exploit the information across regimes. This framework allows breaks in the variance, the regression coefficients or both. Regime duration can be modelled as a Poisson distribution. An new efficient Markov Chain Monte Carlo sampler draws the parameters as one block from the posterior distribution. An application to Canada inflation time series shows the gains in forecasting precision that our model provides. |
| Keywords: | multiple change-points; regime duration; inflation targeting; predictive density; MCMC |
| JEL: | C51 C22 C11 |
| Date: | 2012–02–22 |
| URL: | http://d.repec.org/n?u=RePEc:pra:mprapa:36870&r=for |
| By: | Banerjee, A.; Bystrov, V.; Mizen, P. |
| Abstract: | Much of the literature on interest rate pass through assumes banks set retail rates by observing current market rates. We argue instead that banks anticipate the direction of short-term market rates when setting interest rates on loans, mortgages and deposits. If anticipated rates - captured by forecasts of short-term interest rates or future markets - are important, the empirical specifications of many previous studies that omit them could be misspecified. Including such forecasts requires a detailed consideration of the information in the yield curve and alternative forecasting models. In this paper we use two methods to extract anticipated changes to short-term market rates - a level, slope, curvature model and a principal components model - at many horizons, before including them in a model of retail rate adjustment for four interest rates in four major euro area economies. We find a significant role for forecasts of market rates in determining interest rate pass through; alternative specifications with futures information yield comparable results. We conclude that it is important to include anticipated changes in market rates to avoid misspecification in pass through estimation. |
| Keywords: | forecasting, factor models, interest rates, pass-through. |
| JEL: | C32 C53 E43 E44 |
| Date: | 2012 |
| URL: | http://d.repec.org/n?u=RePEc:bfr:banfra:361&r=for |
| By: | Ola L{\o}vsletten; Martin Rypdal |
| Abstract: | We introduce tools for inference in the multifractal random walk introduced by Bacry et al. (2001). These tools include formulas for smoothing, filtering and volatility forecasting. In addition, we present methods for computing conditional densities for one- and multi-step returns. The inference techniques presented in this paper, including maximum likelihood estimation, are applied to data from the Oslo Stock Exchange, and it is observed that the volatility forecasts based on the multifractal random walk have a much richer structure than the forecasts obtained from a basic stochastic volatility model. |
| Date: | 2012–02 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:1202.5376&r=for |
| By: | Ralf Sabiwalsky |
| Abstract: | Basel II Pillar 3 reports provide information about banks' exposure towards a number of risk factors, such as corporate credit risk and interest rate risk. Previous studies nd that the quality of such information is likely to be weak. We analyze the marginal contribution of pillar 3 exposure data to the quality of equity volatility forecasts for individual banks. Our method uses (local in time) measures of risk factor risk using a multivariate stochastic volatility model for ve risk factors, and uses measures of bank sensitivity with respect to these risk factors. We use two sets of sensitivity measures. One takes into account pillar 3 information, and the other one does not. Generally, we generate volatility forecasts as if no market prices of equity were available for the bank the forecast is made for. We do this for banks for which such data is, in fact, available so that we can conduct ex post - tests of the quality of volatility forecasts. We nd that (1) pillar 3 information allows for a better-than-random ranking of banks according to their risk, but (2) pillar 3 exposure data does not help reduce volatility forecast error magnitude. |
| Keywords: | Risk Reporting, Stochastic Volatility, Risk Factors |
| JEL: | G17 G21 |
| Date: | 2012–02 |
| URL: | http://d.repec.org/n?u=RePEc:hum:wpaper:sfb649dp2012-008&r=for |
| By: | Katsuhiko Muramiya (Research Institute for Economics and Business Administration, Kobe University, Nada-ku, Kobe, Japan); Kazuhisa Otogawa (Graduate School of Business Administration, Kobe University) |
| Abstract: | We use trade size to distinguish between individuals and institutions and then examine their trading behaviors around earnings announcements using data from the Tokyo Stock Exchange. Japanese listed firms have a distinctive financial reporting system in that they report actual earnings for prior and current years, and in addition, almost all of them release management earnings forecasts for the next year. Under this unique setting, we test whether individuals respond differently from institutions to the same earnings news. We document the following results: (1) With regard to current earnings, individuals (institutions) strongly respond to simplistic random walk forecast errors (analyst forecast errors), while do not always respond to analyst forecast errors (simplistic random walk forecast errors). (2) With regard to management earnings forecasts, both individuals and institutions use them, but individuals react to them literally. In contrast to na¨ıve trading by individuals, institutions rationally respond to them with their predicted optimistic bias in mind. Overall, our results suggest that individuals' trading is so na¨ıve as if they use nothing other than the information released at the time of earning announcement, while institutions' trading is so sophisticated. |
| Date: | 2012–02 |
| URL: | http://d.repec.org/n?u=RePEc:kob:dpaper:dp2012-06&r=for |
| By: | Tito Nícias Teixeira da Silva Filho |
| Abstract: | Core inflation is under attack. Empirically, experts have become increasingly disappointed with its actual performance. Theoretically, while some claim that it is a key inflation predictor others argue that, by construction, that cannot be one of its main properties, at least in the short run. Even if true, core inflation could still be useful if it provides good directional inflation forecasts. The evidence presented here using U.S., Canadian and Brazilian data shows that this does not seem to be the case. Directional forecasts are often no better than a coin toss, especially from the level model. The gap model’s forecasts are wrong, on average, at least 20% of the time. More crucially, they are usually no better than a simple moving average of headline inflation. |
| Date: | 2012–01 |
| URL: | http://d.repec.org/n?u=RePEc:bcb:wpaper:266&r=for |
| By: | Bec, F.; Bouabdallah, O.; Ferrara, L. |
| Abstract: | This paper proposes a two-regime Bounce-Back Function augmented Self-Exciting Threshold AutoRegression (SETAR) which allows for various shapes of recoveries from the recession regime. It relies on the bounce-back effects first analyzed in a Markov-Switching setup by Kim, Morley and Piger [2005] and recently extended by Bec, Bouabdallah and Ferrara [2011a]. This approach is then applied to post-1973 quarterly growth rates of French, German, Italian, Spanish and Euro area real GDPs. Both the linear autoregression and the standard SETAR without bounce-back effect null hypotheses are strongly rejected against the Bounce-Back augmented SETAR alternative in all cases but Italy. The relevance of our proposed model is further assessed by the comparison of its short-term forecasting performances with the ones obtained from a linear autoregression and a standard SETAR. It turns out that the bounce-back models one-step ahead forecasts generally outperform the other ones, and particularly so during the last recovery period in 2009Q3-2010Q4. |
| Keywords: | Threshold autoregression, bounce-back effects, asymmetric business cycles. |
| JEL: | E32 C22 |
| Date: | 2012 |
| URL: | http://d.repec.org/n?u=RePEc:bfr:banfra:360&r=for |