Model selection for nonlinear time series

We investigate the finite-sample performance of model selection criteria for local linear regression by simulation. Similarly to linear regression, the penalization term depends on the number of parameters of the model. In the context of nonparametric regression, we use a suitable quantity to account for the Equivalent Number of Parameters as previously suggested in the literature. We consider the following criteria: Rice T, FPE, AIC, Corrected AIC and GCV. To make results comparable with other data-driven selection criteria we consider also Leave-Out CV. We show that the properties of the penalization schemes are very different for some linear and nonlinear models. Finally, we set up a goodness-of-fit test for linearity based on bootstrap methods. The test has correct size and very high power against the alternatives investigated. Application of the methods proposed to macroeconomic and financial time series shows that there is evidence of nonlinearity.First version received: September 2002/Final version received : October 2003I would like to thank Cees Diks, Cars Hommes and an anonymous referee for useful comments that significantly improved the paper.     

A Semiparametric Analysis of the Life Cycle-Permanent Income Hypothesis

This paper considers a semiparametric regression model to test the various implications of the Life Cycle-Permanent Income (LCP) hypothesis proposed by Hall (1978). The semiparametric regression model does not require any parametric assumption on the functional form of the unknown utility function in our analysis. In contrast, the linear regression models frequently used in the literature are justified under specific parametric forms of the utility function and may lead to a misleading conclusion on the LCP hypothesis if the parameterization is incorrect. Using both linear and semiparametric regression models, tests of the martingale property of consumption along with several specification tests are performed on the U. S quarterly data from 1947 to 1990. The results from the semiparametric model do not differ significantly from those from the linear model and suggest some evidences against the implications of the LCP hypothesis. [C14]     

线性回归模型结构稳定性的Eviews诊断

线性回归模型是计量经济学中最基本的模型,当模型涉及时间序列数据时就可能出现模型结构稳定性的问题,即模型参数可能随着时间的变化而变化。本文根据Chow检验的方法步骤。介绍线性回归模型结构稳定性的Eviews诊断方法。     

The relationship between regression and weighted averages in forecasting

The known relationship between linear regression and weighted averages are examined more extensively in the case of forecasting by univariate regression. Several surprising results accrue from the analysis, including the fact that it is possible for a given observation to have no bearing whatsoever on the forecast.     

A small-sample correction for the likelihood ratio test

Anderson's (1958, p. 208) asymptotic expansion of the likelihood ratio criterion for testing linear hypotheses about regression coefficients in the multivariate linear model is used to derive a small-sample correction factor that may be applied in a more general context.     

Applied R in the Classroom

R, with support from a number of add-on packages, is an excellent teaching tool that can greatly enhance learning of exploratory data analysis (EDA) and linear regression. This article illustrates using R along with the packages tidyverse , GGally , esquisse and lindia to walk through an example of basic EDA and linear regression that might be used in an introductory class along with code examples for every step. Teaching with R can give learners control of their analysis, lower the intimidating coding barrier, and provide a platform on which to learn modeling and analysis.     

An examination of the economic significance of stock return predictability in UK stock returns

We explore the out-of-sample performance of domestic UK asset allocation strategies that use forecasts of expected returns from a linear predictive regression and those that are implied by asset pricing models such as the capital asset pricing model (CAPM) or arbitrage pricing theory (APT). Our findings suggest that using forecasts of expected returns from the predictive regression generate significant benefits in out-of-sample performance. We find the performance of the strategies using expected return forecasts implied by the CAPM or APT is lower than the predictive regression strategy. However, with binding investment constraints, the performance of the APT matches that of the predictive regression.     

An encompassing test for non-nested quantile regression models

We propose an encompassing test for non-nested linear quantile regression models and show that it has an asymptotic χ² distribution. It is also shown that the proposed test is a regression rank score test in a comprehensive model under conditional homogeneity. Our simulation results indicate that the proposed test performs very well in finite samples.     

THE EFFECT OF GOVERNMENT SPENDING ON ECONOMIC GROWTH: TESTING THE NON‐LINEAR HYPOTHESIS

Theoretical models suggest a non‐linear relationship between government size and long‐run economic growth. However, testing this hypothesis empirically in cross‐country studies is complicated by the endogeneity of government spending and the accurate identification of inflexion points. This paper examines the non‐linear hypothesis by incorporating threshold analysis in a cross‐country growth regression. The methodology utilizes a sample‐splitting framework and follows an objective strategy for identifying and testing changes in the slope. The results provide evidence in support of the non‐linear hypothesis for a broad panel of countries.     

黄冈市人口增长对经济发展影响的研究

通过近年来黄冈市人口增长与经济发展的现状分析,运用主成分分析法和多元线性回归分析法,建立了黄冈市人口增长对经济发展影响的回归模型,结合实证分析的结果,揭示黄冈市人口增长对其经济发展的影响,为有关经济政策的制定和解决人口问题提供参考。     

An omnibus test for heteroskedasticity

Tests for heteroskedasticity in the classical linear regression framework typically require the researcher to specify the form of variance heterogeneity. This paper proposes a simulation-based test with power to detect unspecified forms of heteroskedasticity.     

Tests in contingency tables as regression tests

Applied researchers often use tests based on contingency tables, especially in preliminary data analysis and diagnostic testing. We show that many such tests may be alternatively implemented by testing for coefficient restrictions in linear regression systems.     

Can interest rate changes help predict future stock price movements? Evidence from the German market

By applying a linear regression model to monthly time series data from the German equity and money markets, this paper challenges the conventional viewpoint that historical data do not possess any explanatory power for future stock market returns.     

江苏省电力消费与产业结构的实证分析--基于多元线形回归分析

文章对江苏2000—2010年产业结构演变中三次产业的电力消费情况进行多元线形回归分析,通过对偏回归系数的计算与分析,提出需要注重产业融合发展,强化产业承载能力等建议。     

基于门限分位点回归的条件VaR风险度量

门限分位点回归模型是线性分位点回归模型的改进,是一种更加客观实际的非线性估计方法。利用该模型实证分析了单只股票(民生银行)的条件VaR。选择一种流动性风险指标作为条件,经过分析发现,由门限分位点回归模型得到结果能够更好地描述市场,也能更好地预测市场风险。     

基于灰色系统理论的旅游客源预测模型--以中国入境旅游客源为例

本文基于灰色系统理论,以中国1978年以来入境客源为例,建构了旅游客源预测的灰色模型,并与常用的线性模型的预测精度进行了定量对比。所得基本结论为:①旅游系统是灰色系统,灰色系统理论是研究旅游现象的有力工具;②中国入境客源灰色GM(1,1)模型的预测精度高于线性模型;③就灰色预测而言,如果原始数据列较长(例如21年),对原始数据进行平滑处理所得预测模型精度较高,而对原始数据进行取自然对数的处理所得模型预测精度较低;如果原始数据列较短(例如5年),对原始数据进行取自然对数的预处理所得模型预测精度较高。本文所得结论对其他旅游序列预测应该具有一定的借鉴与参考价值。     

基于二元回归分析的火灾数据研究

根据国家2003—2007年火灾的相关统计数据,应用回归分析,研究了火灾引起的经济损失与火灾中伤人数目及烧毁建筑面积之间的关系,建立了二元线性回归模型,对方程的精度进行了相关性检验。     

A modification of the CUSUM test in the linear regression model with lagged dependent variables

We consider testing for structural change in a dynamic linear regression model, and show that the well known CUSUM test, which has been initially devised only for the standard static model, can easily be modified such as to remain asymptotically valid also in this nonstandard situation.     

线性回归模型参数线性约束检验的Eviews实施

线性回归模型是计量经济模型的主要形式,其参数除了需要显著性检验和方程显著性检验外,有时还需要对其参数线性约束条件的检验。本文介绍运用Eviews软件对模型参数线性约束的F检验、沃尔德检验与t检验三种方法。     

Interaction effects in econometrics

We provide practical advice for applied economists regarding robust specification and interpretation of linear regression models with interaction terms. We replicate a number of prominently published results using interaction effects and examine if they are robust to reasonable specification permutations.