
on Forecasting 
By:  Richard A. Ashley. 
Keywords:  forecasting,forecast loss functions,stochastic dominance. 
Date:  2006 
URL:  http://d.repec.org/n?u=RePEc:vpi:wpaper:e0610&r=for 
By:  Clements, Michael P (Department of Economics, University of Warwick); Harvey, David I (School of Economics, University of Nottingham) 
Abstract:  We consider tests of forecast encompassing for probability forecasts, for both quadratic and logarithmic scoring rules. We propose test statistics for the null of forecast encompassing, present the limiting distributions of the test statistics, and investigate the impact of estimating the forecasting models’ parameters on these distributions. The smallsample performance of the various statistics is investigated, both in terms of small numbers of forecasts and model estimation sample sizes. Two empirical applications show the usefulness of the tests for the evaluation of recession probability forecasts from logit models with different leading indicators as explanatory variables, and for evaluating surveybased probability forecasts. Probability forecasts ; encompassing tests ; recession probabilities 
JEL:  C12 C15 C53 
Date:  2006 
URL:  http://d.repec.org/n?u=RePEc:wrk:warwec:774&r=for 
By:  Clements, Michael P (Department of Economics, University of Warwick); Galvão, Ana Beatriz (Bank of Portugal) 
Abstract:  Although many macroeconomic series such as US real output growth are sampled quarterly, many potentially useful predictors are observed at a higher frequency. We look at whether a recently developed mixed datafrequency sampling (MIDAS) approach can improve forecasts of output growth and inflation. We carry out a number of related realtime forecast comparisons using various indicators as explanatory variables. We find that MIDAS model forecasts of output growth are more accurate at horizons less than one quarter using coincident indicators ; that MIDAS models are an effective way of combining information from multiple indicators ; and that the forecast accuracy of the unemploymentrate Phillips curve for inflation is enhanced using the MIDAS approach. 
Keywords:  Data frequency ; multiple predictors ; combination ; realtime forecasting 
JEL:  C51 C53 
Date:  2006 
URL:  http://d.repec.org/n?u=RePEc:wrk:warwec:773&r=for 
By:  Sophocles N. Brissimis (Bank of Greece, Economic Research Department and University of Piraeus); Nicholas S. Magginas (National Bank of Greece) 
Abstract:  This paper evaluates the role of inflationforecast heterogeneity in US monetary policy making. The deviation between private and central bank inflation forecasts is identified as a factor increasing inflation persistence and thus calling for a policy reaction. An optimal policy rule is derived by the minimization under discretion of a standard central bank loss function subject to a Phillips curve, modified to include the forecast deviation, and a forwardlooking aggregate demand equation. This rule, which itself includes the forecast deviation as an additional argument, is estimated for the period 19741998, covering the Chairmanships of Arthur Burns, Paul Volcker and Alan Greenspan, by using realtime forecasts of inflation and the output gap obtained from the FOMC’s Greenbook and the Survey of Professional Forecasters. The estimated rule remains remarkably stable over the whole sample period, challenging the conventional view of a structural break following Volcker’s appointment as Chairman of the Fed. Finally, the substantial decline in the significance of the interestrate smoothing term in the rule indicates that monetary policy inertia may, to a large extent, be an artifact of serially correlated inflationforecast errors that feed into policy decisions in real time. 
Keywords:  Forwardlooking model; Monetary policy reaction function; Expectations formation; Inflation expectations 
JEL:  D84 E31 E43 E52 E58 
Date:  2006–03 
URL:  http://d.repec.org/n?u=RePEc:bog:wpaper:35&r=for 
By:  Sophocles N. Brissimis (Bank of Greece, Economic Research Department and University of Piraeus); Nicholas S. Magginas (National Bank of Greece) 
Abstract:  The ability of the New Keynesian Phillips curve to explain US inflation dynamics when official central bank forecasts (Greenbook forecasts) are used as a proxy for inflation expectations is examined. The New Keynesian Phillips curve is estimated on quarterly data spanning the period 1970Q11998Q2 against the alternative of the Hybrid Phillips curve, which allows for a backwardlooking component in the pricesetting behavior in the economy. The results are compared to those obtained using actual data on future inflation as conventionally employed in empirical work under the assumption of rational expectations. The empirical evidence provides, in contrast to most of the relevant literature, considerable support for the standard forwardlooking New Keynesian Phillips curve when inflation expectations are measured using official inflation forecasts. In this case, lagged inflation terms become insignificant in the hybrid specification. The usefulness of real unit labor cost as the preferred proxy for real marginal cost in recent empirical work on the Phillips curve is confirmed by our results. 
Keywords:  Money demand; Inflation; Phillips curve; Real marginal cost; Realtime data; GMM estimation 
JEL:  C13 C52 E31 E37 E50 E52 
Date:  2006–05 
URL:  http://d.repec.org/n?u=RePEc:bog:wpaper:38&r=for 
By:  Takatoshi Ito; Yuko Hashimoto 
Abstract:  This paper examines the price impact and the predictability of the exchange rate movement using the transaction data recorded in the electronic broking system of the spot foreign exchange market. The number of actual deals at the ask (or bid side) for a specified time interval may be regarded as "order flows" to buy (or sell) in Richard Lyons' work. First, the contemporaneous impact of order flows on the quote and deal prices are analyzed. Second, the price predictability is examined. Our forecasting equations of the exchange rate for the next X minutes (X=1, 5, 15, 30) show that coefficients are significantly different from zero in both 5min and 1min forecast horizons, but the significance disappears in the 30minute interval. The tstatistics become larger as the prediction window becomes shorter. Price impacts of deals at one side of the market are significant but shortlived. Market participants, if they can observe and analyze all the transactions information in real time, may be able to extract information to predict the price movements in the following next few minutes. 
JEL:  F31 F33 G15 
Date:  2006–11 
URL:  http://d.repec.org/n?u=RePEc:nbr:nberwo:12682&r=for 
By:  Paul De Grauwe; Pablo Rovira Kaltwasser 
Abstract:  This paper presents a behavioral finance model of the exchange rate. Agents forecast the exchange rate by means of very simple rules. They can choose between three groups of forecasting rules: fundamentalist, extrapolative and momentum rules. Agents using a fundamentalist rule are not able to observe the true value of the fundamental exchange and therefore have to rely on an estimate of this variable to make a forecast. Based on simulation analysis we find that two types of equilibria exist, a fundamental and a nonfundamental one. Both the probability of finding a particular type of equilibrium and the probability of switching between different types of equilibria depend on the number of rules available to agents. Furthermore, we find that the exchange rate dynamics is sensitive to initial conditions and to the risk perception about the underlying fundamental. Both results are dependent on the number of forecasting rules. 
JEL:  C53 F31 
Date:  2006 
URL:  http://d.repec.org/n?u=RePEc:ces:ceswps:_1849&r=for 
By:  Schumacher, Christian; Breitung, Jörg 
Abstract:  This paper discusses a factor model for estimating monthly GDP using a large number of monthly and quarterly time series in realtime. To take into account the different periodicities of the data and missing observations at the end of the sample, the factors are estimated by applying an EM algorithm combined with a principal components estimator. We discuss the insample properties of the estimator in realtime environments and methods for outofsample forecasting. As an empirical application, we estimate monthly German GDP in realtime, discuss the nowcast and forecast accuracy of the model and the role of revisions. Furthermore, we assess the contribution of timely monthly data to the forecast performance. 
Keywords:  monthly GDP, EM algorithm, principal components, factor models 
JEL:  C53 E37 
Date:  2006 
URL:  http://d.repec.org/n?u=RePEc:zbw:bubdp1:5097&r=for 
By:  George A. Christodoulakis (Bank of Greece and Manchester Business School) 
Abstract:  This paper examines the behaviour of the demand for money in Greece during 1976:12000:4, a period that included many of the influences that cause moneydemand instability. Two empirical methodologies, vector error correction (VEC) modelling and secondgeneration random coefficient (RC) modelling, are used to estimate the demand for money. The coefficients of both the VEC and RC procedures support the hypothesis that the demand for money becomes more responsive to both the own rate of return on money balances and the opportunity cost of holding money because of financial deregulation. In general, both procedures also support the hypothesis that the income elasticity of money demand declines over time as a result of technological improvements in the payments system and the development of money substitutes, which lead to economies of scale in holding money. 
Keywords:  Asymmetric Loss Preferences, Forecast Rationality, GDP Growth Forecasts, GMM Estimation, LinLin. 
JEL:  C1 C44 C53 E17 E27 
Date:  2005–12 
URL:  http://d.repec.org/n?u=RePEc:bog:wpaper:30&r=for 
By:  Ralf Brueggemann; Helmut Luetkepohl; Massimiliano Marcellino 
Abstract:  It is investigated whether Euroarea variables can be forecast better based on synthetic time series for the preEuro period or by using just data from Germany for the preEuro period. Our forecast comparison is based on quarterly data for the period 1970Q1  2003Q4 for ten macroeconomic variables. The years 2000  2003 are used as forecasting period. A range of different univariate forecasting methods is applied. Some of them are based on linear autoregressive models and we also use some nonlinear or timevarying coefficient models. It turns out that most variables which have a similar level for Germany and the Euroarea such as prices can be better predicted based on German data while aggregated European data are preferable for forecasting variables which need considerable adjustments in their levels when joining German and EMU data. These results suggest that for variables which have a similar level for Germany and the Euroarea it may be reasonable to consider the German preEMU data for studying economic problems in the Euroarea. 
Keywords:  Aggregation, forecasting, European monetary union, constructing EMU data 
JEL:  C22 C53 
Date:  2006 
URL:  http://d.repec.org/n?u=RePEc:eui:euiwps:eco2006/30&r=for 
By:  Eleni Constantinou (Department of Accounting and Finance, The Philips College, 46 Lamias Street, CY2100, Nicosia,); Robert Georgiades (Department of Accounting and Finance, The Philips College, 46 Lamias Street, CY2100, Nicosia,); Avo Kazandjian (Department of Business Studies, The Philips College, 46 Lamias Street, CY2100, Nicosia, Cyprus.); George Kouretas (Department of Economics, University of Crete, Greece) 
Abstract:  This paper provides an analysis of regime switching in volatility and outofsample forecasting of the Cyprus Stock Exchange using daily data for the period 19962002. We first model volatility regime switching within a univariate MarkovSwitching framework. Modelling stock returns within this context can be motivated by the fact that the change in regime should be considered as a random event and not predictable. The results show that linearity is rejected in favour of a MS specification, which forms statistically an adequate representation of the data. Two regimes are implied by the model; the high volatility regime and the low volatility one and they provide quite accurately the state of volatility associated with the presence of a rational bubble in the capital market of Cyprus. Another implication is that there is evidence of regime clustering. We then provide outofsample forecasts of the CSE daily returns using two competing nonlinear models, the univariate Markov Switching model and the Artificial Neural Network Model. The comparison of the outofsample forecasts is done on the basis of forecast accuracy, using the Diebold and Mariano (1995) test and forecast encompassing, using the Clements and Hendry (1998) test. The results suggest that both nonlinear models equivalent in forecasting accuracy and forecasting encompassing and therefore on forecasting performance. 
Keywords:  Regime switching, artificial neural networks, stock returns, forecast 
JEL:  G 
Date:  2005–01 
URL:  http://d.repec.org/n?u=RePEc:crt:wpaper:0502&r=for 
By:  Clements, Michael P (Department of Economics, University of Warwick) 
Abstract:  We ask whether the different types of forecasts made by individual survey respondents are mutually consistent, using the SPF survey data. We compare the point forecasts and central tendencies of probability distributions matched by individual respondent, and compare the forecast probabilities of declines in output with the probabilities implied by the probability distributions. When the expected associations between these different types of forecasts do not hold for some idividuals, we consider whether the discrepancies we observe are consistent with rational behaviour by agents with asymmetric loss functions. 
Keywords:  Rationality ; probability forecasts ; probability distributions 
JEL:  C53 E32 E37 
Date:  2006 
URL:  http://d.repec.org/n?u=RePEc:wrk:warwec:772&r=for 
By:  Christoph Hartz (University of Munich); Stefan Mittnik (University of Munich, Center for Financial Studies and ifo); Marc S. Paolella (University of Zurich) 
Abstract:  A resampling method based on the bootstrap and a biascorrection step is developed for improving the ValueatRisk (VaR) forecasting ability of the normalGARCH model. Compared to the use of more sophisticated GARCH models, the new method is fast, easy to implement, numerically reliable, and, except for having to choose a window length L for the biascorrection step, fully data driven. The results for several different financial asset returns over a long outofsample forecasting period, as well as use of simulated data, strongly support use of the new method, and the performance is not sensitive to the choice of L. 
Keywords:  Bootstrap, GARCH, ValueatRisk 
JEL:  C22 C53 C63 G12 
Date:  2006–11–03 
URL:  http://d.repec.org/n?u=RePEc:cfs:cfswop:wp2000623&r=for 
By:  De Mol, Christine; Giannone, Domenico; Reichlin, Lucrezia 
Abstract:  This paper considers Bayesian regression with normal and doubleexponential priors as forecasting methods based on large panels of time series. We show that, empirically, these forecasts are highly correlated with principal component forecasts and that they perform equally well for a wide range of prior choices. Moreover, we study the asymptotic properties of the Bayesian regression under Gaussian prior under the assumption that data are quasi collinear to establish a criterion for setting parameters in a large crosssection. 
Keywords:  Bayesian VAR, ridge regression, Lasso regression, principal components, large crosssections 
JEL:  C11 C13 C33 C53 
Date:  2006 
URL:  http://d.repec.org/n?u=RePEc:zbw:bubdp1:5040&r=for 
By:  Daniel J. Benjamin; Jesse M. Shapiro 
Abstract:  We showed 10second, silent video clips of unfamiliar gubernatorial debates to a group of experimental participants and asked them to predict the election outcomes. The participants' predictions explain more than 20 percent of the variation in the actual twoparty vote share across the 58 elections in our study, and their importance survives a range of controls, including state fixed effects. In a horse race of alternative forecasting models, participants' visual forecasts significantly outperform economic variables in predicting vote shares, and are comparable in predictive power to a measure of incumbency status. Adding policy information to the video clips by turning on the sound tends, if anything, to worsen participants' accuracy, suggesting that naïveté may be an asset in some forecasting tasks. 
JEL:  D72 J45 
Date:  2006–11 
URL:  http://d.repec.org/n?u=RePEc:nbr:nberwo:12660&r=for 
By:  Faidon Kalfaoglou (Bank of Greece); Alexandros Sarris (Food and Agricultural Organization of the United Nations and University of Athens) 
Abstract:  This paper evaluates the role of inflationforecast heterogeneity in US monetary policy making. The deviation between private and central bank inflation forecasts is identified as a factor increasing inflation persistence and thus calling for a policy reaction. An optimal policy rule is derived by the minimization under discretion of a standard central bank loss function subject to a Phillips curve, modified to include the forecast deviation, and a forwardlooking aggregate demand equation. This rule, which itself includes the forecast deviation as an additional argument, is estimated for the period 19741998, covering the Chairmanships of Arthur Burns, Paul Volcker and Alan Greenspan, by using realtime forecasts of inflation and the output gap obtained from the FOMC’s Greenbook and the Survey of Professional Forecasters. The estimated rule remains remarkably stable over the whole sample period, challenging the conventional view of a structural break following Volcker’s appointment as Chairman of the Fed. Finally, the substantial decline in the significance of the interestrate smoothing term in the rule indicates that monetary policy inertia may, to a large extent, be an artifact of serially correlated inflationforecast errors that feed into policy decisions in real time. 
Keywords:  Market discipline, transparency, bank risk 
JEL:  G18 G21 G28 
Date:  2006–04 
URL:  http://d.repec.org/n?u=RePEc:bog:wpaper:36&r=for 