http://repub.eur.nl/res/concept/jel-C53/ Recent publications classified by JEL Code C53 en http://repub.eur.nl/static-eur/img/logo.png http://repub.eur.nl http://repub.eur.nl/res/pub/39848/ 2013-04-19T00:00:00Z This paper develops a testing framework for comparing the predictive accuracy of copula-based multivariate density forecasts, focusing on a specific part of the joint distribution. The test is framed in the context of the Kullback-Leibler Information Criterion, but using (out-of-sample) conditional likelihood and censored likelihood in order to focus the evaluation on the region of interest. Monte Carlo simulations document that the resulting test statistics have satisfactory size and power properties in small samples. In an empirical application to daily exchange rate returns we find evidence that the dependence structure varies with the sign and magnitude of returns, such that different parametric copula models achieve superior forecasting performance in different regions of the support. Our analysis highlights the importance of allowing for lower and upper tail dependence for accurate forecasting of common extreme appreciation and depreciation of different currencies. Diks, C.G.H. Panchenko, V. Sokolinskiy, O. Dijk, D.J.C. van http://repub.eur.nl/res/pub/39847/ 2013-04-15T00:00:00Z Accurate prediction of risk measures such as Value at Risk (VaR) and Expected Shortfall (ES) requires precise estimation of the tail of the predictive distribution. Two novel concepts are introduced that offer a specific focus on this part of the predictive density: the censored posterior, a posterior in which the likelihood is replaced by the censored likelihood; and the censored predictive likelihood, which is used for Bayesian Model Averaging. We perform extensive experiments involving simulated and empirical data. Our results show the ability of these new approaches to outperform the standard posterior and traditional Bayesian Model Averaging techniques in applications of Value-at-Risk prediction in GARCH models. Gatarek, L.T. Hoogerheide, L.F. Honing, K. http://repub.eur.nl/res/pub/39841/ 2013-04-11T00:00:00Z It is common practice to evaluate fixed-event forecast revisions in macroeconomics by regressing current forecast revisions on one-period lagged forecast revisions. Under weak-form (forecast) efficiency, the correlation between the current and one-period lagged revisions should be zero. The empirical findings in the literature suggest that this null hypothesis of zero correlation is rejected frequently, where the correlation can be either positive (which is widely interpreted in the literature as “smoothing”) or negative (which is widely interpreted as “over-reacting”). We propose a methodology to interpret such non-zero correlations in a straightforward and clear manner. Our approach is based on the assumption that numerical forecasts can be decomposed into both an econometric model and random expert intuition. We show that the interpretation of the sign of the correlation between the current and one-period lagged revisions depends on the process governing intuition, and the current and lagged correlations between intuition and news (or shocks to the numerical forecasts). It follows that the estimated non-zero correlation cannot be given a direct interpretation in terms of smoothing or over-reaction. Chang, C.L. Bruijn, B. de Franses, Ph.H.B.F. McAleer, M.J. http://repub.eur.nl/res/pub/39840/ 2013-04-08T00:00:00Z This paper presents the Matlab package DeCo (Density Combination) which is based on the paper by Billio et al. (2013) where a constructive Bayesian approach is presented for combining predictive densities originating from different models or other sources of information. The combination weights are time-varying and may depend on past predictive forecasting performances and other learning mechanisms. The core algorithm is the function DeCo which applies banks of parallel Sequential Monte Carlo algorithms to filter the time-varying combination weights. The DeCo procedure has been implemented both for standard CPU computing and for Graphical Process Unit (GPU) parallel computing. For the GPU implementation we use the Matlab parallel computing toolbox and show how to use General Purposes GPU computing almost effortless. This GPU implementation comes with a speed up of the execution time up to seventy times compared to a standard CPU Matlab implementation on a multicore CPU. We show the use of the package and the computational gain of the GPU version, through some simulation experiments and empirical applications. Casarin, R. Grassi, S. Ravazzolo, F. Dijk, H.K. van http://repub.eur.nl/res/pub/39434/ 2013-03-25T00:00:00Z Many macroeconomic forecasts and forecast updates like those from IMF and OECD typically involve both a model component, which is replicable, as well as intuition, which is non-replicable. Intuition is expert knowledge possessed by a forecaster. If forecast updates are progressive, forecast updates should become more accurate, on average, as the actual value is approached. Otherwise, forecast updates would be neutral. The paper proposes a methodology to test whether macroeconomic forecast updates are progressive, where the interaction between model and intuition is explicitly taken into account. The data set for the empirical analysis is for Taiwan, where we have three decades of quarterly data available of forecasts and their updates of the inflation rate and real GDP growth rate. Our empirical results suggest that the forecast updates for Taiwan are progressive, and that progress can be explained predominantly by improved intuition. Chang, C.L. Franses, Ph.H.B.F. McAleer, M.J. http://repub.eur.nl/res/pub/38198/ 2012-10-29T00:00:00Z We propose a Bayesian combination approach for multivariate predictive densities which relies upon a distributional state space representation of the combination weights. Several specifications of multivariate time-varying weights are introduced with a particular focus on weight dynamics driven by the past performance of the predictive densities and the use of learning mechanisms. In the proposed approach the model set can be incomplete, meaning that all models can be individually misspecified. A Sequential Monte Carlo method is proposed to approximate the filtering and predictive densities. The combination approach is assessed using statistical and utility-based performance measures for evaluating density forecasts. Simulation results indicate that, for a set of linear autoregressive models, the combination strategy is successful in selecting, with probability close to one, the true model when the model set is complete and it is able to detect parameter instability when the model set includes the true model that has generated subsamples of data. For the macro series we find that incompleteness of the models is relatively large in the 70's, the beginning of the 80's and during the recent financial crisis, and lower during the Great Moderation. With respect to returns of the S&P 500 series, we find that an investment strategy using a combination of predictions from professional forecasters and from a white noise model puts more weight on the white noise model in the beginning of the 90's and switches to giving more weight to the professional forecasts over time. Billio, M. Casarin, R. Ravazzolo, F. Dijk, H.K. van http://repub.eur.nl/res/pub/37538/ 2012-10-25T00:00:00Z We introduce a heuristic bias-adjustment for the transaction price-based realized range estimator of daily volatility in the presence of bid-ask bounce and non-trading. The adjustment is an extension of the estimator proposed in Christensen et al. (2009). We relax the assumption that all intra-day high (low) transaction prices are at the ask (bid) quote. Using data-based simulations we obtain estimates of the probability that a given intraday range is (upward or downward) biased or not, which we use for a more refined bias-adjustment of the realized range estimator. Both Monte Carlo simulations and an empirical application involving a liquid and a relatively illiquid S&P500 constituent demonstrate that ex post measures and ex ante forecasts based on the heuristically adjusted realized range compare favorably to existing bias-adjusted (two time scales) realized range and (two time scales) realized variance estimators. Bannouh, K. Martens, M.P.E. Dijk, D.J.C. van http://repub.eur.nl/res/pub/30685/ 2011-12-15T00:00:00Z We propose a new and simple methodology to estimate the loss function associated with experts' forecasts. Under the assumption of conditional normality of the data and the forecast distribution, the asymmetry parameter of the lin-lin and linex loss function can easily be estimated using a linear regression. This regression also provides an estimate for potential systematic bias in the forecasts of the expert. The residuals of the regression are the input for a test for the validity of the normality assumption. We apply our approach to a large data set of SKU-level sales forecasts made by experts and we compare the outcomes with those for statistical model-based forecasts of the same sales data. We find substantial evidence for asymmetry in the loss functions of the experts, with underprediction penalized more than overprediction. Franses, Ph.H.B.F. Legerstee, R. Paap, R. http://repub.eur.nl/res/pub/30684/ 2011-11-30T00:00:00Z We propose a multivariate combination approach to prediction based on a distributional state space representation of the weights belonging to a set of Bayesian predictive densities which have been obtained from alternative models. Several specifications of multivariate time-varying weights are introduced with a particular focus on weight dynamics driven by the past performance of the predictive densities and the use of learning mechanisms. In the proposed approach the model set can be incomplete, meaning that all models are individually misspecified. The approach is assessed using statistical and utility-based performance measures for evaluating density forecasts of US macroeconomic time series and surveys of stock market prices. For the macro series we find that incompleteness of the models is relatively large in the 70's, the beginning of the 80's and during the recent financial crisis; structural changes like the Great Moderation are empirically identified by our model combination and the predicted probabilities of recession accurately compare with the NBER business cycle dating. Model weights have substantial uncertainty attached and neglecting this may seriously affect results. With respect to returns of the S&P 500 series, we find that an investment strategy using a combination of predictions from professional forecasters and from a white noise model puts more weight on the white noise model in the beginning of the 90's and switches to giving more weight to the left tail of the professional forecasts during the start of the financial crisis around 2008. Billio, M. Casarin, R. Ravazzolo, F. Dijk, H.K. van http://repub.eur.nl/res/pub/26867/ 2011-11-14T00:00:00Z This paper deals with the analysis and evaluation of sales forecasts of managers, given that it is unknown how they constructed their forecasts. Our goal is to find out whether these forecasts are rational. To examine deviations from rationality, we argue that one has to approximate how the managers could have generated the forecasts. We describe several ways to construct these approximate expressions. The analysis of a large set of a single manager's forecasts for sales of pharmaceutical products illustrates the practical usefulness of our methodology. Bruijn, B. de Franses, Ph.H.B.F. http://repub.eur.nl/res/pub/26526/ 2011-09-30T00:00:00Z Experts can rely on statistical model forecasts when creating their own forecasts. Usually it is not known what experts actually do. In this paper we focus on three questions, which we try to answer given the availability of expert forecasts and model forecasts. First, is the expert forecast related to the model forecast and how? Second, how is this potential relation influenced by other factors? Third, how does this relation influence forecast accuracy? We propose a new and innovative two-level Hierarchical Bayes model to answer these questions. We apply our proposed methodology to a large data set of forecasts and realizations of SKU-level sales data from a pharmaceutical company. We find that expert forecasts can depend on model forecasts in a variety of ways. Average sales levels, sales volatility, and the forecast horizon influence this dependence. We also demonstrate that theoretical implications of expert behavior on forecast accuracy are reflected in the empirical data. Legerstee, R. Franses, Ph.H.B.F. http://repub.eur.nl/res/pub/26506/ 2011-09-26T00:00:00Z We analyze the behavior of experts who quote forecasts for monthly SKU-level sales data where we compare data before and after the moment that experts received different kinds of feedback on their behavior. We have data for 21 experts located in as many countries who make SKU-level forecasts for a variety of pharmaceutical products for October 2006 to September 2007. We study the behavior of the experts by comparing their forecasts with those from an automated statistical program, and we report the forecast accuracy over these 12 months. In September 2007 these experts were given feedback on their behavior and they received a training at the headquarters' office, where specific attention was given to the ins and outs of the statistical program. Next, we study the behavior of the experts for the 3 months after the training session, that is, October 2007 to December 2007. Our main conclusion is that in the second period the experts' forecasts deviated lesser from the statistical forecasts and that their accuracy improved substantially. Legerstee, R. Franses, Ph.H.B.F. http://repub.eur.nl/res/pub/26086/ 2011-09-02T00:00:00Z This paper develops a novel approach to modeling and forecasting realized volatility (RV) measures based on copula functions. Copula-based time series models can capture relevant characteristics of volatility such as nonlinear dynamics and long-memory type behavior in a flexible yet parsimonious way. In an empirical application to daily volatility for S&P500 index futures, we find that the copula-based RV (C-RV) model outperforms conventional forecasting approaches for one-day ahead volatility forecasts in terms of accuracy and efficiency. Among the copula specifications considered, the Gumbel C-RV model achieves the best forecast performance, which highlights the importance of asymmetry and upper tail dependence for modeling volatility dynamics. Although we find substantial variation in the copula parameter estimates over time, conditional copulas do not improve the accuracy of volatility forecasts. Sokolinskiy, O. Dijk, D.J.C. van http://repub.eur.nl/res/pub/26505/ 2011-09-01T00:00:00Z Patton and Timmermann (2011, 'Forecast Rationality Tests Based on Multi-Horizon Bounds', Journal of Business & Economic Statistics, forthcoming) propose a set of useful tests for forecast rationality or optimality under squared error loss, including an easily implemented test based on a regression that only involves (long-horizon and short-horizon) forecasts and no observations on the target variable. We propose an extension, a simulation-based procedure that takes into account the presence of errors in parameter estimates. This procedure can also be applied in the field of 'backtesting' models for Value-at-Risk. Applications to simple AR and ARCH time series models show that its power in detecting certain misspecifications is larger than the power of well-known tests for correct Unconditional Coverage and Conditional Coverage. Hoogerheide, L.F. Ravazzolo, F. Dijk, H.K. van http://repub.eur.nl/res/pub/26508/ 2011-09-01T00:00:00Z Kernel ridge regression is gaining popularity as a data-rich nonlinear forecasting tool, which is applicable in many different contexts. This paper investigates the influence of the choice of kernel and the setting of tuning parameters on forecast accuracy. We review several popular kernels, including polynomial kernels, the Gaussian kernel, and the Sinc kernel. We interpret the latter two kernels in terms of their smoothing properties, and we relate the tuning parameters associated to all these kernels to smoothness measures of the prediction function and to the signal-to-noise ratio. Based on these interpretations, we provide guidelines for selecting the tuning parameters from small grids using cross-validation. A Monte Carlo study confirms the practical usefulness of these rules of thumb. Finally, the flexible and smooth functional forms provided by the Gaussian and Sinc kernels makes them widely applicable, and we recommend their use instead of the pop ular polynomial kernels in general settings, in which no information on the data-generating process is available. Exterkate, P. http://repub.eur.nl/res/pub/25712/ 2011-07-25T00:00:00Z Factor construction methods are widely used to summarize a large panel of variables by means of a relatively small number of representative factors. We propose a novel factor construction procedure that enjoys the properties of robustness to outliers and of sparsity; that is, having relatively few nonzero factor loadings. Compared to more traditional factor construction methods, we find that this procedure leads to better interpretable factors and to a favorable forecasting performance, both in a Monte Carlo experiment and in two empirical applications to large data sets, one from macroeconomics and one from microeconomics. Croux, C. Exterkate, P. http://repub.eur.nl/res/pub/25708/ 2011-07-13T00:00:00Z This paper examines which macroeconomic and financial variables are most informative for the federal funds target rate decisions made by the Federal Open Market Committee (FOMC) from a forecasting perspective. The analysis is conducted for the FOMC decision during the period January 1990 - June 2008, using dynamic ordered probit models with a Bayesian endogenous variable selection methodology and real-time data for a set of 33 candidate predictor variables. We find that indicators of economic activity and forward-looking term structure variables as well as survey measures have most predictive ability. For the full sample period, in-sample probability forecasts achieve a hitrate of 90 percent. Based on out-of-sample forecasts for the period January 2001 - June 2008, 82 percent of the FOMC decisions are predicted correctly. Hauwe, S. van den Dijk, D.J.C. van Paap, R. http://repub.eur.nl/res/pub/25610/ 2011-07-01T00:00:00Z In McAleer et al. (2010b), a robust risk management strategy to the Global Financial Crisis (GFC) was proposed under the Basel II Accord by selecting a Value-at-Risk (VaR) forecast that combines the forecasts of different VaR models. The robust forecast was based on the median of the point VaR forecasts of a set of conditional volatility models. In this paper we provide further evidence on the suitability of the median as a GFC-robust strategy by using an additional set of new extreme value forecasting models and by extending the sample period for comparison. These extreme value models include DPOT and Conditional EVT. Such models might be expected to be useful in explaining financial data, especially in the presence of extreme shocks that arise during a GFC. Our empirical results confirm that the median remains GFC-robust even in the presence of these new extreme value models. This is illustrated by using the S&P500 index before, during and after the 2008-09 GFC. We investigate the performance of a variety of single and combined VaR forecasts in terms of daily capital requirements and violation penalties under the Basel II Accord, as well as other criteria, including several tests for independence of the violations. The strategy based on the median, or more generally, on combined forecasts of single models, is straightforward to incorporate into existing computer software packages that are used by banks and other financial institutions. Santos, P.A. Jimenez-Martin, J-A. McAleer, M.J. Perez-Amaral, T. http://repub.eur.nl/res/pub/25614/ 2011-07-01T00:00:00Z It is well known that the Basel II Accord requires banks and other Authorized Deposit-taking Institutions (ADIs) to communicate their daily risk forecasts to the appropriate monetary authorities at the beginning of each trading day, using one or more risk models, whether individually or as combinations, to measure Value-at-Risk (VaR). The risk estimates of these models are used to determine capital requirements and associated capital costs of ADIs, depending in part on the number of previous violations, whereby realised losses exceed the estimated VaR. McAleer et al. (2009) proposed a new approach to model selection for predicting VaR, consisting of combining alternative risk models, and comparing conservative and aggressive strategies for choosing between VaR models. This paper addresses the question of risk management of risk, namely VaR of VIX futures prices, and extends the approaches given in McAleer et al. (2009) and Chang et al. (2011) to examine how different risk management strategies performed during the 2008-09 global financial crisis (GFC). The empirical results suggest that an aggressive strategy of choosing the Supremum of single model forecasts, as compared with Bayesian and non-Bayesian combinations of models, is preferred to other alternatives, and is robust during the GFC. However, this strategy implies relatively high numbers of violations and accumulated losses, which are admissible under the Basel II Accord. Casarin, R. Chang, C.L. Jimenez-Martin, J-A. McAleer, M.J. Perez-Amaral, T. http://repub.eur.nl/res/pub/23785/ 2011-06-30T00:00:00Z It is common practice to evaluate fixed-event forecast revisions in macroeconomics by regressing current revisions on one-period lagged revisions. Under weak-form efficiency, the correlation between the current and one-period lagged revisions should be zero. The empirical findings in the literature suggest that the null hypothesis of zero correlation between the current and one-period lagged revisions is rejected quite frequently, where the correlation can be either positive or negative. In this paper we propose a methodology to be able to interpret such non-zero correlations in a straightforward manner. Our approach is based on the assumption that forecasts can be decomposed into both an econometric model and expert intuition. The interpretation of the sign of the correlation between the current and one-period lagged revisions depends on the process governing intuition, and the correlation between intuition and news. Franses, Ph.H.B.F. Chang, C.L. McAleer, M.J.