|
on Forecasting |
| By: | Mehmet Balcilar (Department of Economics, Eastern Mediterranean University); Nico Katzke (Department of Economics, University of Stellenbosch); Rangan Gupta (Department of Economics, University of Pretoria) |
| Abstract: | In this paper we test whether the key metals prices of gold and platinum significantly improve inflation forecasts for the South African economy. We also test whether controlling for conditional correlations in a dynamic setup, using bivariate Bayesian-Dynamic Conditional Correlation (B-DCC) models, improves inflation forecasts. To achieve this we compare out-of-sample forecast estimates of the B-DCC model to Random Walk, Autoregressive and Bayesian VAR models. We find that for both the BVAR and BDCC models, improving point forecasts of the Autoregressive model of inflation remains an elusive exercise. This, we argue, is of less importance relative to the more informative density forecasts. For this we find improved forecasts of inflation for the B-DCC models at all forecasting horizons tested. We thus conclude that including metals price series as inputs to inflation models leads to improved density forecasts, while controlling for the dynamic relationship between the included price series and inflation similarly leads to significantly improved density forecasts. |
| Keywords: | Bayesian VAR, Dynamic Conditional Correlation, Density forecasting, Random Walk, Autoregressive model |
| JEL: | C11 C15 E17 |
| Date: | 2015 |
| URL: | http://d.repec.org/n?u=RePEc:sza:wpaper:wpapers235&r=for |
| By: | Dovern, Jonas; Huber, Florian |
| Abstract: | We present evidence that global vectorautoregressive (GVAR) models produce significantly more accurate recession forecasts than country-specific time-series models in a Bayesian framework. This result holds for most countries and forecast horizons as well as for several country groups. |
| Keywords: | GVAR; recession forecast; QPS; probability forecast |
| Date: | 2015–03–17 |
| URL: | http://d.repec.org/n?u=RePEc:awi:wpaper:0585&r=for |
| By: | Roland Döhrn; Philipp an de Meulen |
| Abstract: | In periods of unusual weather, forecasters face a problem of interpreting economic data: Which part goes back to the underlying economic trend and which part arises from a special weather effect? In this paper, we discuss ways to disentangle weather-related from business cycle-related influences on economic indicators. We find a significant influence of weather variables at least on a number of monthly indicators. Controlling for weather effects within these indicators should thus create opportunities to increase the accuracy of indicator-based forecasts. Focusing on quarterly GDP growth in Germany, we find that the accuracy of the RWI short term forecasting model improves but advances are small and not significant. |
| Keywords: | Weather; short term forecasting; bridge equations; forecast accuracy |
| JEL: | C53 E37 |
| Date: | 2015–02 |
| URL: | http://d.repec.org/n?u=RePEc:rwi:repape:0539&r=for |
| By: | Christopher Gibbs (School of Economics, UNSW Business School, UNSW) |
| Abstract: | This paper introduces the concept of a Forecast Combination Equilibrium to model boundedly rational agents who combine a menu of different forecasts using insights from the forecasting literature to mimic the behavior of actual forecasters. The equilibrium concept is consistent with rational expectations under certain conditions, while also permitting multiple, distinct, self-fulfilling equilibria, many of which are stable under least squares learning. The equilibrium concept is applied to a simple Lucas-type monetary model where agents engage in constant gain learning. The combination of multiple equilibria and learning is sufficient to replicate some key features of in ation data, such as time-varying volatility and periodic bouts of high in ation or deflation in a model that experiences only i.i.d. random shocks. |
| Keywords: | Forecast Combination, Adaptive Learning, Expectations, Dynamic Predictor Selection, Inflation, Forecast Combination Puzzle |
| JEL: | E17 E31 C52 C53 |
| Date: | 2015–02 |
| URL: | http://d.repec.org/n?u=RePEc:swe:wpaper:2015-05&r=for |
| By: | Alpaslan YARAR (Selcuk University); Mustafa ONÜÇYILDIZ (Selcuk University); Nuri PEKÇETİN (Metropolitan Municipility of Adana) |
| Abstract: | To development and management of the water resources, fluctuations over the amount of water resources should be determined. Fluctuations depend on the rainfall, runoff, geological, meteorological properties of the area and many others. Scientist studied to determine this process using physical models. But, in recent years Adaptive Neuro Fuzzy Inference System (ANFIS) has being widely used. Even absence of some data to determine these hydrological processes, ANFIS model can be used efficiently. Many data-driven models, including linear, nonparametric or nonlinear approaches, are developed for hydrologic discharge time series prediction in the past decades (Marques et al., 2006). Generally, the prediction techniques for a dynamic system can be roughly divided into two approaches: local and global. Local approach uses only nearby states to make predictions whereas global approach involves all the states. K-Nearest-Neighbors (KNN) algorithm, Artificial Neural Networks (ANN) and Support Vectors Machine (SVM) are some typical forecast methods for dynamic systems (Sivapragasam et al., 2001; Laio et al., 2003; Wang et al., 2006). Kazeminezhad et al. (2005) used an adaptive network-based fuzzy inference systems (ANFIS) model, which is a fuzzy inference system, whose rules parameters are tuned by ANNs, in prediction of wave parameters in fetch-limited condition. Zanganeh et al. (2006) combined GAs and ANFIS models in the problem of prediction of wave parameters.In this study, 5 Flow Observation Station (FOS) in the West Mediterranean Basin in Turkey was modeled to forecast the monthly flow data using ANFIS. It was seen that ANFIS model can be used to forecast the monthly flow efficiently. |
| Keywords: | Forecasting ,Runoff, ANFIS |
| JEL: | C53 C45 C67 |
| Date: | 2014–06 |
| URL: | http://d.repec.org/n?u=RePEc:sek:iacpro:0201599&r=for |
| By: | Josh R. Stillwagon (Department of Economics, Trinity College) |
| Abstract: | This paper offers and tests a unique explanation for the exchange rate determination puzzle. It is not that exchange rates are unrelated to fundamentals, but rather when fundamentals undergo persistent changes it becomes important to measure their effect in terms of how they change relative to what was expected. This result is demonstrated with a simple present discounted value model of the exchange rate and then tested for four USD exchange rates using interest rate forecast data from nearly 50 major banks. Using the polynomially cointegrated VAR, or I(2) CVAR, the interest rate forecast errors are found to have a large and statistically significant impact on the exchange rate even independent of the level and change in the relative interest rate (with t-values in the double digits for all four samples). Further, this effect is greater in the samples with stronger evidence of persistent changes in the interest rate differential. |
| Keywords: | Exchange Rates, Determination Puzzle, Survey Data, Forecast Errors, I(2) Cointegration |
| JEL: | F31 G12 G15 |
| Date: | 2015–02 |
| URL: | http://d.repec.org/n?u=RePEc:tri:wpaper:1501&r=for |
| By: | Edward L. Glaeser; Charles G. Nathanson |
| Abstract: | A modest approximation by homebuyers leads house prices to display three features that are present in the data but usually missing from perfectly rational models: momentum at one-year horizons, mean reversion at five-year horizons, and excess longer-term volatility relative to fundamentals. Valuing a house involves forecasting the current and future demand to live in the surrounding area. Buyers forecast using past transaction prices. Approximating buyers do not adjust for the expectations of past buyers, and instead assume that past prices reflect only contemporaneous demand, as with a capitalization rate formula. Consistent with survey evidence, this approximation leads buyers to expect increases in the market value of their homes after recent house price increases, to fail to anticipate the price busts that follow booms, and to be overconfident in their assessments of the housing market. |
| JEL: | D03 G02 R21 |
| Date: | 2015–03 |
| URL: | http://d.repec.org/n?u=RePEc:nbr:nberwo:21037&r=for |
| By: | Joshua C.C. Chan (Research School of Economics, and Centre for Applied Macroeconomic Analysis, Australian National University); Rodney Strachan (School of Economics, and Centre for Applied Macroeconomic Analysis, University of Queensland; The Rimini Centre for Economic Analysis, Italy) |
| Abstract: | The time-varying parameter vector autoregressive (TVP-VAR) model has been used to successfully model interest rates and other variables. As many short interest rates are now near their zero lower bound (ZLB), a feature not included in the standard TVP-VAR specification, this model is no longer appropriate. However, there remain good reasons to include short interest rates in macro models, such as to study the effect of a credit shock. We propose a TVP-VAR that accounts for the ZLB and study algorithms for computing this model that are less computationally burdensome than others yet handle many states well. To illustrate the proposed approach, we investigate the effect of the zero lower bound of interest rate on transmission of a monetary shock. |
| Date: | 2014–12 |
| URL: | http://d.repec.org/n?u=RePEc:rim:rimwps:42_14&r=for |
| By: | Rahul Mukherjee (Indian Institute of Management Calcutta) |
| Abstract: | Predicting a future observation on the basis of the existing observations is a problem of compelling practical interest in many fields of study including economics and sociology. Bayesian predictive densities, obtained via a prior specification on the underlying population, are commonly used for this purpose. This may, however, induce subjectivity because the resulting predictive set depends on the choice of prior. Moreover, one can as well consider direct frequentist methods which do not require any prior specification. This can again entail results differing from what Bayesian predictive densities yield. Thus there is a need to reconcile all these approaches.The present article aims at addressing this problem. Specifically, we explore predictive sets which have frequentist as well as Bayesian validity for arbitrary priors in an asymptotic sense. Our tools include a connection with locally unbiased tests and a shrinkage argument for Bayesian asymptotics. Our findings apply to general multiparameter statistical models and represent a significant advance over the existing work in this area which caters only to models with a single unknown parameter and that too under certain restrictions. Illustrative examples are given. Computation and simulation studies show that our results work very well in finite samples. |
| Keywords: | Asymptotic theory, locally unbiased test, posterior predictive density, shrinkage argument |
| JEL: | C11 C15 |
| Date: | 2014–12 |
| URL: | http://d.repec.org/n?u=RePEc:sek:iacpro:0901514&r=for |