|
on Econometric Time Series |
| By: | Korobilis, Dimitris |
| Abstract: | Machine learning methods are becoming increasingly popular in economics, due to the increased availability of large datasets. In this paper I evaluate a recently proposed algorithm called Generalized Approximate Message Passing (GAMP) , which has been very popular in signal processing and compressive sensing. I show how this algorithm can be combined with Bayesian hierarchical shrinkage priors typically used in economic forecasting, resulting in computationally efficient schemes for estimating high-dimensional regression models. Using Monte Carlo simulations I establish that in certain scenarios GAMP can achieve estimation accuracy comparable to traditional Markov chain Monte Carlo methods, at a tiny fraction of the computing time. In a forecasting exercise involving a large set of orthogonal macroeconomic predictors, I show that Bayesian shrinkage estimators based on GAMP perform very well compared to a large set of alternatives. |
| Keywords: | high-dimensional inference; compressive sensing; belief propagation; Bayesian shrinkage; dynamic factor models |
| Date: | 2017–01 |
| URL: | http://d.repec.org/n?u=RePEc:esy:uefcwp:19565&r=ets |
| By: | Antonakakis, Nikolaos; Gabauer, David |
| Abstract: | In this study, we propose refined measures of dynamic connectedness based on a TVP-VAR approach, that overcomes certain shortcomings of the connectedness measures introduced originally by Diebold and Yilmaz (2009, 2012, 2014). We illustrate the advantages of the TVP-VAR-based connectedness approach with an empirical analysis on exchange rate volatility connectedness. |
| Keywords: | Dynamic connectedness; TVP-VAR; Exchange rate volatility |
| JEL: | C32 C50 F31 G15 |
| Date: | 2017–04–12 |
| URL: | http://d.repec.org/n?u=RePEc:pra:mprapa:78282&r=ets |
| By: | Francesco Bartolucci (Universita' di Perugia); Claudia Pigini (Universita' Politecnica delle Marche, Dipartimento di Scienze Economiche e Sociali) |
| Abstract: | Strict exogeneity of covariates other than the lagged dependent variable, and conditional on unobserved heterogeneity, is often required for consistent estimation of binary panel data models. This assumption is likely to be violated in practice because of feedback e ects from the past of the outcome variable on the present value of covariates and no general solution is yet available. In this paper, we provide the conditions for a logit model formulation that takes into account feedback e ects without specifying a joint parametric model for the outcome and predetermined explanatory variables. Our formulation is based on the equivalence between Granger's de nition of noncausality and a modi cation of the Sims' strict exogeneity assumption for nonlinear panel data models, introduced by Chamberlain (1982) and for which we provide a more general theorem. We further propose estimating the model parameters with a recent xed-e ects approach based on pseudo conditional inference, adapted to the present case, thereby taking care of the correlation between individual permanent unobserved heterogeneity and the model's covariates as well. Our results hold for short panels with a large number of cross-section units, a case of great interest in microeconomic applications. |
| Keywords: | binary panel data, fixed e ects, feedback e ects, pseudo-conditional inference |
| JEL: | C12 C23 C25 |
| Date: | 2017–04 |
| URL: | http://d.repec.org/n?u=RePEc:anc:wpaper:421&r=ets |
| By: | Xuehai Zhang (Paderborn University); Yuanhua Feng (Paderborn University); Christian Peitz (Paderborn University) |
| Abstract: | The paper proposes a wide class of semiparametric GARCH models by introducing a scale function into a GARCH class model for featuring long-run volatility dynamics, which can be thought of as an MEM (multiplicative error model) with a varying scale function. Our focus is to estimate the scale function under suitable weak moment conditions by means of the Box-Cox transformation of the absolute returns. The estimation of the scale function is independent of any GARCH specification. To overcome the drawbacks of the kernel and the local linear approaches, a non-negatively constrained local linear estimator of the scale function is considered, which is then proposed to fit a suitable parametric GARCH model to the standardized residuals, is used. Asymptotic properties of the proposed nonpara- metric and parametric estimators are studied in detail and iterative plug-in algorithms are developed for selecting the bandwidth and transformation parameters, which are selected by MLE and JB statistic. The algorithms are also carried out independently without any parametric specification in the stationary part. Application to real data sets show that the proposals work very well in practice. |
| Date: | 2017–04 |
| URL: | http://d.repec.org/n?u=RePEc:pdn:ciepap:104&r=ets |
| By: | Yuanhua Feng (Paderborn University); Thomas Gries (Paderborn University) |
| Abstract: | The main purpose of this paper is the development of iterative plug-in algorithms for local polynomial estimation of the trend and its derivatives in macroeconomic time series. In particular, a data-driven lag-window estimator for the variance factor is proposed so that the bandwidth is selected without any parametric assumption on the stationary errors. Further analysis of the residuals using an ARMA model is discussed briefl y. Moreover, confidence bounds for the trend and its derivatives are conducted using some asymptotically unbiased estimates and applied to test possible linearity of the trend. These graphical tools also provide us further detailed features about the economic development. Practical performance of the proposals is illustrated by quarterly US and UK GDP data. |
| Keywords: | Macroeconomic time series, semiparametric modelling, nonparametric regression with dependent errors, bandwidth selection, misspecification test |
| Date: | 2017–04 |
| URL: | http://d.repec.org/n?u=RePEc:pdn:ciepap:102&r=ets |
| By: | Syed Ali Asad Rizvi; Stephen J. Roberts; Michael A. Osborne; Favour Nyikosa |
| Abstract: | In this paper we use Gaussian Process (GP) regression to propose a novel approach for predicting volatility of financial returns by forecasting the envelopes of the time series. We provide a direct comparison of their performance to traditional approaches such as GARCH. We compare the forecasting power of three approaches: GP regression on the absolute and squared returns; regression on the envelope of the returns and the absolute returns; and regression on the envelope of the negative and positive returns separately. We use a maximum a posteriori estimate with a Gaussian prior to determine our hyperparameters. We also test the effect of hyperparameter updating at each forecasting step. We use our approaches to forecast out-of-sample volatility of four currency pairs over a 2 year period, at half-hourly intervals. From three kernels, we select the kernel giving the best performance for our data. We use two published accuracy measures and four statistical loss functions to evaluate the forecasting ability of GARCH vs GPs. In mean squared error the GP's perform 20% better than a random walk model, and 50% better than GARCH for the same data. |
| Date: | 2017–05 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:1705.00891&r=ets |