|
on Forecasting |
| By: | Croonenboreck, Carsten; Ambach, Daniel |
| Abstract: | In the wind power industry, wind speed forecasts are obtained and transformed into wind power forecasts. The Mycielski algorithm has proven to be an accurate predictor for wind speed in short-term scenarios. Moreover, Mycielski has the capability of forecasting wind power directly, instead of wind speed. This article compares wind power forecasts calculated by the Mycielski algorithm to state-of-the-art forecasters. As such, we use the Wind Power Prediction Tool (WPPT) and the recently developed generalization of it, GWPPT (Generalized WPPT). Furthermore, we evaluate statistical time series models such as autoregressive and vector autoregressive models. As an additional benchmark we use the persistence model, which is often used to assess forecasting accuracy. Each model is evaluated and we give a recommendation for the best forecasting model. -- |
| Keywords: | Mycielski algorithm,WPPT,GWPPT,Wind Power,Wind Energy,Forecasting,Prediction |
| JEL: | C35 E27 Q47 |
| Date: | 2014 |
| URL: | http://d.repec.org/n?u=RePEc:zbw:euvwdp:355&r=for |
| By: | Minford, Patrick (Cardiff Business School); Xu, Yongden; Zhou, Peng (Cardiff Business School) |
| Abstract: | Out-of-sample forecasting tests of DSGE models against time-series benchmarks such as an unrestricted VAR are increasingly used to check a) the specification b) the forecasting capacity of these models. We carry out a Monte Carlo experiment on a widely-used DSGE model to investigate the power of these tests. We find that in specification testing they have weak power relative to an in-sample indirect inference test; this implies that a DSGE model may be badly mis-specified and still improve forecasts from an unrestricted VAR. In testing forecasting capacity they also have quite weak power, particularly on the lefthand tail. By contrast a model that passes an indirect inference test of specification will almost definitely also improve on VAR forecasts. |
| Keywords: | Out of sample forecasts; DSGE; VAR; specification tests; indirect inference; forecast performance |
| JEL: | E10 E17 |
| Date: | 2014–07 |
| URL: | http://d.repec.org/n?u=RePEc:cdf:wpaper:2014/11&r=for |
| By: | Kholodilin, Konstantin A.; Thomas, Tobias; Ulbricht, Dirk |
| Abstract: | In an uncertain world, decisions by market participants are based on expectations. Thus, sentiment indicators reflecting expectations are proven at predicting economic variables. However, survey respondents largely perceive the world through media reports. Typically, crude media information, like word-count indices, is used in the prediction of macroeconomic and financial variables. Here, we employ a rich data set provided by Media Tenor International, based on sentiment analysis of opinion-leading media in Germany from 2001 to 2014, transformed into several monthly indices. German industrial production is predicted in a real-time out-of-sample forecasting experiment using more than 17,000 models formed of all possible combinations with a maximum of 3 out of 48 macroeconomic, survey, and media indicators. Media data are indispensable for the prediction of German industrial production both for individual models and as a part of combined forecasts, particularly during the global financial crisis. -- |
| Keywords: | forecast combination,media data,German industrial production,reliability index,R-word |
| JEL: | C10 C52 C53 E32 |
| Date: | 2014 |
| URL: | http://d.repec.org/n?u=RePEc:zbw:dicedp:149&r=for |
| By: | Luba Petersen (Simon Fraser University) |
| Abstract: | This paper explores the importance of accessible and focal information in influencing beliefs and attention in a learning-to-forecast laboratory experiment where subjects are incentivized to form accurate expectations about inflation and the output gap. We consider the effects of salient and accessible forecast error information and learning on subjects' forecasting accuracy and heuristics, and on aggregate stability. Experimental evidence indicates that, while there is considerable heterogeneity in heuristics used, subjects' forecasts can be best described by a constant-gain learning model where subjects respond to forecast errors. Salient forecast error information reduces subjects' overreaction to their errors and leads to greater forecast accuracy, coordination of expectations and macroeconomic stability. The benefits of this focal information are short-lived and diminish with learning. |
| Keywords: | experimental macroeconomics, laboratory experiment, monetary policy, expectations, learning to forecast, availability heuristic, focal points, communication, rational inattention |
| JEL: | C92 E2 E52 D50 D91 |
| Date: | 2014–07 |
| URL: | http://d.repec.org/n?u=RePEc:sfu:sfudps:dp14-05&r=for |
| By: | Hans Christian Müller-Dröge; Tara M. Sinclair; H.O. Stekler |
| Abstract: | In this paper we present an evaluation of forecasts of a vector of variables of the German economy made by different institutions. Our method permits one to evaluate the forecasts for each year and then if one is interested to combine the years. We use our method to determine an overall winner for a forecasting competition across twenty-five different institutions for a single time period using a vector of eight key economic variables. Typically forecasting competitions are judged on a variable-by-variable basis, but our methodology allows us to determine how each competitor performed overall. We find that the Bundesbank was the overall winner for 2013. |
| Keywords: | Mahalanobis Distance, forecasting competition, GDP components, German macroeconomic data |
| JEL: | C5 E2 E3 |
| Date: | 2014–07 |
| URL: | http://d.repec.org/n?u=RePEc:een:camaaa:2014-55&r=for |
| By: | Ortiz-Arango, Francisco; Cabrera-Llanos, Agustín I.; Venegas-Martínez, Francisco |
| Abstract: | This paper is aimed at developing a new kind of non-parametrical artificial neural network useful to forecast exchange rates. To do this, we departure from the so-called Differential or Dynamic neural Networks (DNN) and extend the tracking procedure. Under this approach, we examine the daily closing values of the exchange rates of the Euro against the US dollar, the Japanese yen and the British pound. With our proposal, Extended DNN or EDNN, we perform the tracking procedure from February 15, 1999, to August 31, 2013, and, subsequently, the forecasting procedure from September 2 to September 13, 2013. The accuracy of the obtained results is remarkable, since the percentage of the error in the predicted values is within the range from 0.001% to 0.69% in the forecasting period. |
| Keywords: | Exchange rates, artificial neural network, differential neural network, tracking and forecasting. |
| JEL: | G17 |
| Date: | 2014–08–01 |
| URL: | http://d.repec.org/n?u=RePEc:pra:mprapa:57720&r=for |
| By: | Christian Gouriéroux (CREST and University of Toronto); Joann Jasiak (York University) |
| Abstract: | This paper revisits the filtering and prediction in noncausal and mixed autoregressive processes and provides a simple alternative set of methods that are valid for processes with infinite variances. The prediction method provides complete predictive densities and prediction intervals at any finite horizon H, for univariate and multivariate processes. It is based on an unobserved component representation of noncausal processes. The filtering procedure for the unobserved components is provided along with a simple back-forecasting estimator for the parameters of noncausal and mixed models and a simulation algorithm for noncausal and mixed autoregressive processes. The approach is illustrated by simulations |
| Keywords: | Noncausal Process, Nonlinear Prediction, Filtering, Look-Ahead Estimator, Speculative Bubble, Technical Analysis |
| JEL: | C14 G32 G23 |
| Date: | 2014–04 |
| URL: | http://d.repec.org/n?u=RePEc:crs:wpaper:2014-15&r=for |
| By: | Maciej Krzak; Grzegorz Poniatowski; Katarzyna W¹sik |
| Abstract: | This paper presents forecasts for the Financial Stress Index (FSI) and the Economic Sensitivity Index (ESI) for the period 2014-2015 for six countries in the region, namely the Czech Republic, Estonia, Hungary, Latvia, Lithuania and Poland. It is a continuation of the endeavor to construct synthetic indices measuring financial stress and economic sensitivity for twelve Central and East European countries using the Principal Component Analysis. In order to obtain forecasts of the FSI, we estimated Vector Autoregression (VAR) models on monthly data for the period 2001-2012 separately for all the countries. Using quarterly historical values of ESI and FSI, we estimated Dynamic Panel Data Model for the complete sample of countries. Parameters of the model were later used for forecasting the ESI. Obtained results suggest that the FSI will start to rise in 2014 in the Czech Republic, Lithuania, and Estonia. For Latvia and Hungary, we observed a conversion in the trend, i.e. at the beginning of 2015, when the index should start to fall. According to our forecasts, the ESI will be rising in the next two years, except for Hungary, where we predict a continuous decrease in economic sensitivity. |
| Keywords: | financial stress, economic sensitivity, economic indicators, Central and Eastern Europe |
| JEL: | G01 E32 C43 |
| Date: | 2014–07 |
| URL: | http://d.repec.org/n?u=RePEc:sec:cnstan:0474&r=for |
| By: | F. Pancotto; G. Pignataro; D. Raggi |
| Abstract: | This paper investigates the role of higher order beliefs in the formation of exchange rates. Our model combines a standard macroeconomic dynamics for the exchange rates with a microeconomic specification of agents' heterogeneity and their interactions. The empirical analysis relies on a state space model estimated through Bayesian methods. We exploit data on macroeconomic fundamentals in a panel of subjective forecasts on the euro/dollar exchange rate. The equilibrium strategy on the optimization process of the predictors shows that higher order beliefs is the relevant factor in performing individual forecasting. Moreover public information, namely past exchange rates and fundamentals, plays a crucial role as a coordination device to generate expectations among agents on the basis of their forecasting abilities. |
| JEL: | D82 F41 C11 |
| Date: | 2014–07 |
| URL: | http://d.repec.org/n?u=RePEc:bol:bodewp:wp957&r=for |
| By: | Paresh Kumar Narayan (Deakin University); Huson Ali Ahmed (Deakin University) |
| Abstract: | In this paper our goal is to examine the importance of skewness in decision making, in particular on investor utility. We use time-series daily data on sectoral stock returns on the Indian stock exchange. We test for sectoral stock return predictability using commonly used financial ratios, namely, the book-to-market, dividend yield and price-earnings ratio. We find strong evidence of predictability. Using this evidence of predictability, we forecast sectoral stock returns for each of the sectors in our sample, allowing us to devise trading strategies that account for skewness of returns. We discover evidence that accounting for skewness leads not only to higher utility compared to a model that ignores skewness, but utility is sector-dependent. |
| Keywords: | Returns; Skewness; Predictability; Utility; Investor. |
| Date: | 2014–07–07 |
| URL: | http://d.repec.org/n?u=RePEc:dkn:ecomet:fe_2014_11&r=for |
| By: | Rafi, Umar; Masih, Mansur |
| Abstract: | This paper analyzes the Shariah compliant nature of Islamic banks (IB) by using Time Series Multivariate Forecasting techniques to test the correlation and direction of causality between interest rates and IB financing . Islamic finance defines a 0% Interest rate, both on the asset and on the liability side. Thus, in a utopian Islamic financial system, any movement in interest rates should have no direct impact on any aspect of any Islamic bank. However, the supposition of IBs being genuinely Shariah compliant from a Credit Risk perspective has been challenged by many Shariah scholars. Using Malaysia as a test case, this paper measures changes in KLIBOR (Kuala Lampur Interbank Offer Rate) and tests them for correlations and directional causality with the IB Lending rate (used as a proxy measure for financing by Malaysian IBs). If a correlation and causality can be established between KLIBOR and financing by IBs, then it is an indication that IB’s may not be genuinely Shariah compliant. This research is original in that it attempts to relate an important issue of a fiqhi nature to data analysis, via some time series forecasting techniques. It also discusses the policy impacts of the results, and the subsequent risk faced by the Regulators in managing the Interest rate risks for a financial system structured on dual banking - Islamic and Conventional. The findings of the research tend to indicate a correlation and lead-lag causality relationship between Interest rate changes and Islamic bank financing. |
| Keywords: | Islamic banks, Shariah compliant, Time series techniques, Malaysia |
| JEL: | C22 C58 G21 G28 |
| Date: | 2014–07–27 |
| URL: | http://d.repec.org/n?u=RePEc:pra:mprapa:57711&r=for |
| By: | James Morley |
| Abstract: | The presence of "economic slack" directly implies that an economy can grow quickly without any necessary offsetting slow growth or retrenchment in the future. Based on this link between economic slack and future economic growth, I argue for a forecastbased estimate of the output gap as a measure of economic slack. This approach has the advantage of being robust to different assumptions about the underlying structure of the economy and allows for empirical analysis of a Phillips Curve relationship between the output gap and inflation. I apply this approach to investigate economic slack and its relationship with inflation for selected economies in Asia and the Pacific, taking into account structural breaks in long-run growth and uncertainty about the appropriate forecasting model. The estimated output gap is highly asymmetric for most the economies and implies a convex Phillips Curve in many of the cases. |
| Keywords: | output gap; model averaging; business cycle asymmetry; convex Phillips Curve |
| Date: | 2014–06 |
| URL: | http://d.repec.org/n?u=RePEc:bis:biswps:451&r=for |
| By: | Arnak S. Dalalyan (CREST-ENSAE); Mohamed Hebiri (Université Paris Est); Johannes Lederer (Cornell University) |
| Abstract: | Although the Lasso has been extensively studied, the relationship between its prediction performance and the correlations of the covariates is not fully understood. In this paper, we give new insights into this relationship in the context of multiple linear regression. We show, in particular, that the incorporation of a simple correlation measure into the tuning parameter leads to a nearly optimal prediction performance of the Lasso even for highly correlated covariates. However, we also reveal that for moderately correlated covariates, the prediction performance of the Lasso can be mediocre irrespective of the choice of the tuning parameter. For the illustration of our approach with an important application, we deduce nearly optimal rates for the least-squares estimator with total variation penalty |
| Keywords: | multiple linear regression, sparse recovery, total variation penalty, oracle inequalities |
| Date: | 2014–02 |
| URL: | http://d.repec.org/n?u=RePEc:crs:wpaper:2014-05&r=for |