|
on Econometric Time Series |
| By: | Mingxuan Song (Vrije Universiteit Amsterdam); Bernhard van der Sluis (Erasmus University Rotterdam); Yicong Lin (Vrije Universiteit Amsterdam and Tinbergen Institute) |
| Abstract: | Time-varying regression models with trends are commonly used to analyze long-term tendencies and evolving relationships in data. However, statistical inference for parameter paths is challenging, and recent literature has proposed various bootstrap methods to address this issue. Despite this, no software package in any language has yet offered the recently developed tools for conducting inference in time-varying regression models. We propose PyTimeVar, a Python package that implements nonparametric estimation along with multiple new bootstrap-assisted inference methods. It provides a range of bootstrap techniques for constructing pointwise confidence intervals and simultaneous bands for parameter curves. Additionally, the package includes four widely used methods for modeling trends and time-varying relationships. This allows users to compare different approaches within a unified environment. |
| Keywords: | time-varying, bootstrap, nonparametric estimation, boosted Hodrick-Prescott filter, power-law trend, score-driven, state-space |
| JEL: | C14 C22 C87 |
| Date: | 2024–11–03 |
| URL: | https://d.repec.org/n?u=RePEc:tin:wpaper:20240060 |
| By: | Francisco Blasques (Vrije Universiteit Amsterdam and Tinbergen Institute); Janneke van Brummelen (Vrije Universiteit Amsterdam and Tinbergen Institute); Paolo Gorgi (Vrije Universiteit Amsterdam and Tinbergen Institute); Siem Jan Koopman (Vrije Universiteit Amsterdam and Tinbergen Institute) |
| Abstract: | We introduce a nonlinear semi-parametric model that allows for the robust filtering of a common stochastic trend in a multivariate system of cointegrated time series. The observation-driven stochastic trend can be specified using flexible updating mechanisms. The model provides a general approach to obtain an outlier-robust trend-cycle decomposition in a cointegrated multivariate process. A simple two-stage procedure for the estimation of the parameters of the model is proposed. In the first stage, the loadings of the common trend are estimated via ordinary least squares. In the second stage, the other parameters are estimated via Gaussian quasi-maximum likelihood. We formally derive the theory for the consistency of the estimators in both stages and show that the observation-driven stochastic trend can also be consistently estimated. A simulation study illustrates how such robust methodology can enhance the filtering accuracy of the trend compared to a linear approach as considered in previous literature. The practical relevance of the method is shown by means of an application to spot prices of oil-related commodities. |
| Keywords: | consistency, cycle, non-stationary time series, two-step estimation, vector autoregression |
| JEL: | C13 C32 |
| Date: | 2024–11–03 |
| URL: | https://d.repec.org/n?u=RePEc:tin:wpaper:20240062 |
| By: | Xiaolei Wang (University of Melbourne); Tomasz Wo\'zniak (University of Melbourne) |
| Abstract: | The R package bsvarSIGNs implements state-of-the-art algorithms for the Bayesian analysis of Structural Vector Autoregressions identified by sign, zero, and narrative restrictions. It offers fast and efficient estimation thanks to the deployment of frontier econometric and numerical techniques and algorithms written in C++. The core model is based on a flexible Vector Autoregression with estimated hyper-parameters of the Minnesota prior and the dummy observation priors. The structural model can be identified by sign, zero, and narrative restrictions, including a novel solution, making it possible to use the three types of restrictions at once. The package facilitates predictive and structural analyses using impulse responses, forecast error variance and historical decompositions, forecasting and conditional forecasting, as well as analyses of structural shocks and fitted values. All this is complemented by colourful plots, user-friendly summary functions, and comprehensive documentation. The package was granted the Di Cook Open-Source Statistical Software Award by the Statistical Society of Australia in 2024. |
| Date: | 2025–01 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2501.16711 |
| By: | Stéphane Surprenant |
| Abstract: | Recent rises in macroeconomic volatility have prompted the introduction of quantile vector autoregression (QVAR) models to forecast macroeconomic risk. This paper provides an extensive evaluation of the predictive performance of QVAR models in a pseudo-out-of-sample experiment spanning 112 monthly US variables over 40 years, with horizons of 1 to 12 months. We compare QVAR with three parametric benchmarks: a Gaussian VAR, a generalized autoregressive conditional heteroskedasticity VAR and a VAR with stochastic volatility. QVAR frequently, significantly and quantitatively improves upon the benchmarks and almost never performs significantly worse. Forecasting improvements are concentrated in the labour market and interest and exchange rates. Augmenting the QVAR model with factors estimated by principal components or quantile factors significantly enhances macroeconomic risk forecasting in some cases, mostly in the labour market. Generally, QVAR and the augmented models perform equally well. We conclude that both are adequate tools for modeling macroeconomic risks. |
| Keywords: | Econometrics and statistical methods; Business fluctuations and cycles |
| JEL: | C53 E37 C55 |
| Date: | 2025–01 |
| URL: | https://d.repec.org/n?u=RePEc:bca:bocawp:25-4 |